The Global Drug Facility as an intervention in the market for tuberculosis drugs

Abstract Objective To investigate funding for the Global Drug Facility since 2001 and to analyse the facility’s influence on the price of high-quality tuberculosis drugs. Methods Data on the price of tuberculosis drugs were obtained from the Global Drug Facility for 2001 to 2012 and, for the private sector in 15 countries, from IMS Health for 2002 to 2012. Data on funding of the facility were also collected. Findings Quality-assured tuberculosis drugs supplied by the Global Drug Facility were generally priced lower than drugs purchased in the private sector. In 2012, just three manufacturers accounted for 29.9 million United Stated dollars (US$) of US$ 44.5 million by value of first-line drugs supplied. The Global Fund to Fight AIDS, Tuberculosis and Malaria provided 73% (US$ 32.5 million of US$ 44.5 million) and 89% (US$ 57.8 million of US $65.2 million) of funds for first- and second-line drugs, respectively. Between 2010 and 2012, the facility’s market share of second-line tuberculosis drugs increased from 26.1% to 42.9%, while prices decreased by as much as 24% (from US$ 1231 to US$ 939). Conversely, the facility’s market share of first-line drugs fell from 37.2% to 19.2% during this time, while prices increased from US$ 9.53 to US$ 10.2. Conclusion The price of tuberculosis drugs supplied through the facility was generally less than that on the private market. However, to realize its full potential and meet the needs of more tuberculosis patients, the facility requires more diverse and stable public funding and greater flexibility to participate in the private market.


Introduction
Tuberculosis remains a global public health concern. In 2013, there were an estimated 9 million incident cases worldwide, 480 000 of which involved multidrug-resistant tuberculosis. 1 For tuberculosis as well as other conditions, disease control depends on more than the existence of curative treatment -it also depends on the drug supply, which is ultimately mediated by the pharmaceutical market. [2][3][4][5][6][7] Consequently, disease control is profoundly influenced by the functioning of this market, particularly in resource-poor settings with a high disease burden. In addition, despite the existence of international quality-assurance standards, tuberculosis drugs are often either substandard or counterfeit. [8][9][10] The use of substandard drugs reduces the chance of successful treatment and promotes the emergence of drug-resistance. 11 Although the patents have expired on many tuberculosis drugs, the power of individual low-income countries with a high disease burden to negotiate cheaper treatment is limited. Second-line treatment for multidrug-resistant tuberculosis involves more protracted and complex chemotherapy and can cost a hundred times more than treating drug-sensitive tuberculosis. 12,13 In light of these issues, the Global Drug Facility was launched by the Stop TB Partnership in 2001 with the aim of using donor funding to consolidate demand from different countries and negotiate lower prices for quality-assured tuberculosis drugs. 14,15 The facility now occupies a unique position in the global market for these drugs -in 2011, it supplied enough drugs to treat 35% of publicly notified cases of tuberculosis worldwide and an estimated 24% of all incident cases. 16 However, the facility is only one participant in a complex, global tuberculosis drugs market (Fig. 1). Other drug purchasers include those in the private sector, national tuberculosis programmes and, in certain cases, donors themselves. In this environment, a defining feature of the Global Drug Facility model is the central role that international quality-assurance standards play in its operation: they are embedded in overall quality management so that stringent public procurement standards can be met. 17 In the absence of such a framework, even manufacturers concerned about quality may find that the benefits of acquiring international quality-assurance certification do not necessarily outweigh the investments needed to meet these standards. By creating a large, stable market, a mechanism such as the Global Drug Facility provides clear incentives for a supply of drugs that meet international quality-assurance standards. In 2012, the value of this market for tuberculosis drugs exceeded 109 million United States dollars (US$).
Given that the Global Drug Facility plays such a large role in the tuberculosis drugs market, it is important to have some understanding of its influence on both sales volumes and drug prices. The aim of this study was to investigate changes in the price of the tuberculosis treatments supplied by the Global Drug Facility over the past 12 years of its operation and changes in its funding. In addition, we compared the price of tuberculosis treatment supplied by the Global Drug Facility with that of equivalent drugs purchased on the private market in 15 countries.
Objective To investigate funding for the Global Drug Facility since 2001 and to analyse the facility's influence on the price of high-quality tuberculosis drugs. Methods Data on the price of tuberculosis drugs were obtained from the Global Drug Facility for 2001 to 2012 and, for the private sector in 15 countries, from IMS Health for 2002 to 2012. Data on funding of the facility were also collected. Findings Quality-assured tuberculosis drugs supplied by the Global Drug Facility were generally priced lower than drugs purchased in the private sector. In 2012, just three manufacturers accounted for 29.9 million United Stated dollars (US$) of US$ 44.5 million by value of first-line drugs supplied. The Global Fund to Fight AIDS, Tuberculosis and Malaria provided 73% (US$ 32.5 million of US$ 44.5 million) and 89% (US$ 57.8 million of US $65.2 million) of funds for first-and second-line drugs, respectively. Between 2010 and 2012, the facility's market share of second-line tuberculosis drugs increased from 26.1% to 42.9%, while prices decreased by as much as 24% (from US$ 1231 to US$ 939). Conversely, the facility's market share of first-line drugs fell from 37.2% to 19.2% during this time, while prices increased from US$ 9.53 to US$ 10.2. Conclusion The price of tuberculosis drugs supplied through the facility was generally less than that on the private market. However, to realize its full potential and meet the needs of more tuberculosis patients, the facility requires more diverse and stable public funding and greater flexibility to participate in the private market.
The Global Drug Facility and the tuberculosis drugs market Nimalan Arinaminpathy et al.

Methods
The main funding flows in the global tuberculosis drug market are shown in Fig. 1, in which the solid lines indicate the flows for which data were available for our study. Details of the value of the funding channels labelled A in the figure were obtained from procurement data from the Global Drug Facility for the period 2001 to 2012. We derived the number of courses of treatment supplied from these data as described previously. 16 In calculating drug prices, we incorporated the combination of drugs used in a full course of treatment for a single patient (Table 1). For second-line treatment, to cover a wide range of possible treatment regimens, we considered a cheaper, low-end regimen and a more expensive, high-end regimen (Table 1), as in previous work. 16 We used Global Drug Facility data to calculate the cost of a single standard unit of treatment: (i) a fixed-dose combination pill for first-line treatment; and (ii) a pill or vial of injectable compound for second-line treatment. We then derived the cost of a course of treatment for an individual patient using the number of standard units required, as shown in Table 1. Generally we used the mean unit price for each drug and therefore the mean price of each treatment course but we also considered the price range by using the maximum and minimum unit prices for each drug. All prices are expressed in US$, the currency in which the Global Drug Facility purchases and supplies drugs.
In Fig. 1, funding channels B and C represent the private market. Data on these channels were obtained for 2002 to 2012 from IMS Health -an organization that collects information on drug purchases in a range of countries. Data from IMS Health covered 15 countries, including 10 with a high burden of tuberculosis and 11 with a high burden of multidrug-resistant tuberculosis ( Table 2). These countries represented the range of support received from the Global Drug Facility: for example, India has been a major purchaser of drugs through the facility in recent years, whereas South Africa has had almost no involvement. We calculated the price of a treatment course as described above. To achieve consistency with Global Drug Facility data, we converted prices expressed in other currencies into US$ using the exchange rates in force at the time of each transaction.  a Although 2011 treatment guidelines for multidrug-resistant tuberculosis increased the recommended treatment duration for injectable drugs from 6 to 8 months, orders delivered in 2012 had all been placed under previous guidelines. Accordingly, we assumed that all patients were treated for 6 months. b Low-end, second-line treatment regimens were those at the lower end of the price range for all possible regimens. c High-end, second-line treatment regimens were those at the more expensive end of the price range for all possible regimens.

Research
The Global Drug Facility and the tuberculosis drugs market Nimalan Arinaminpathy et al.
For this study, the private market included all sources of tuberculosis drugs that were not supplied by the Global Drug Facility or through any other international financing mechanism, irrespective of whether the drugs were purchased by public or private sector organizations (i.e. channels B and C in Fig. 1). We did not consider other public sources of drugs (i.e. channel D in Fig. 1) because of a lack of systematic price data. Since IMS Health data come from a variety of sources (e.g. retailers and hospitals), incorporate different taxes (e.g. sales and import taxes) and may include discounts for large purchase volumes, it was difficult to compare prices directly. Accordingly, we compared ex-works prices -that is, the prices of drugs purchased and collected at the site of their manufacture. For the private market, we used IMS Health estimates of ex-works prices; for drugs supplied by the Global Drug Facility, we used ex-works prices from facility purchasing data. It was not possible to quantify the uncertainty in IMS Health estimates of ex-works prices because relevant data were not available. To address this limitation, we estimated the magnitude of the price bias that would be needed to negate the findings of our analysis. We adjusted all prices for in-flation in each country separately using data on consumer price indices from the World Bank. Then, to investigate global trends, we averaged prices across countries, weighted by the quantity of drugs supplied to each country.
Finally, for channel C in Fig. 1, it was not possible to compare countries, as it was for channel B, because of a lack of systematic, public data on the price of drugs procured by national tuberculosis programmes directly from manufacturers. One exception was South Africa, which has published procurement data for its tuberculosis programme. 18 In this case, we were able to make a comparison with the Global Drug Facility's prices. Fig. 3 show the change in donor involvement with the Global Drug Facility between 2007 and 2012 for firstand second-line tuberculosis drugs, respectively. Fig. 4, Fig. 5, Fig. 6 and Fig. 7 show the corresponding involvement of selected recipient countries and manufacturers with the facility. One key change in that period was a reduction in bilateral funding from the United Kingdom's Department for International Development for first-line tuberculosis drugs in India. As a result, India stopped receiving these drugs through the Global Drug Facility. Overall, the proportion of the Global Drug Facility's funding that came from the Global Fund to Fight AIDS, Tuberculosis and Malaria increased over time: in 2012, it was 73% (US$ 32.5 million of US$ 44.5 million) and 89% (US$ 57.8 million of US$ 65.2 million) for first-and second-line drugs, respectively. On the supply side, manufacturing remained highly concentrated: the largest three manufacturers together accounted for more than 67% by value of the first-line drugs supplied ($29.9 million of $44.5 million).

Drug price dynamics
In our analysis, we looked at the prices paid for treatment by national tuberculosis programmes supplied by the Global Drug Facility rather than the bid prices initially put forward by manufacturers. Fig. 9  In addition to the mean prices shown in Fig. 9, Fig. 12 shows minimum and maximum prices globally between 2002 and 2012. As might be expected, given that a central purchasing entity was being compared with a diverse private market, the variation in Global Drug Facility prices was markedly less than the variation in private market prices. Fig. 13 and Fig. 14 illustrate the variation between 2002 and 2012 in the price of a course of treatment with first-and second-line drugs, respectively, in selected countries. It shows that the price of drugs supplied by the Global Drug Facility was less than that of drugs available in the private market for all countries. Fig. 15 and Fig. 16 (both available at: http://www.who. int/bulletin/volumes/93/4/14-147256) display the price of individual first-and second-line treatments, respectively, obtained through the Global Drug Facility relative to that of treatment purchased from the private market between 2002 and 2012. The price of most drugs was consistently higher when purchased from the private market. The exceptions were protionamide, capreomycin and kanamycin -their mean price on the private market was 33% (US$ 0.020 versus US$ 0.062), 44% (US$ 1.24 versus US$ 2.84) and 11% (US$ 0.10 versus US$ 0.97) respectively, of the corresponding price from the Global Drug Facility. Nonetheless, since kanamycin accounts for only around 20% (US$ 189 of US$ 939), of the price of a course of low-end, second-line treatment from the Global Drug Facility the overall price of treatment was still lower than it would have been on the private market. Fig. 17 shows the ratio of the price of tuberculosis drugs procured directly  from manufacturers by the national tuberculosis programme in South Africa to the price of drugs from the Global Drug Facility. Again the figure illustrates that, with the exception of kanamycin, the price of drugs supplied by the Global Drug Facility was lower than that of drugs obtained directly from private markets. Moreover, it should be noted that, although manufacturers supplied the Global Drug Facility with drugs that met international quality-assurance standards, many had different production lines that were used to supply other clients, including national programmes. Overall therefore, drugs, including kanamycin, that were supplied by sources other than the Global Drug Facility were of uncertain quality, whether or not they were provided by manufacturers who also supplied the facility.

Discussion
Our analysis suggests that a mechanism such as the Global Drug Facility can indeed secure lower prices for drugs that meet international quality-assurance standards than are available for unregulated drugs of unknown quality on the private market. Moreover, the Global Drug Facility's prices varied considerably less than those in the private market. This could greatly assist planning, both for countries procuring drugs and for manufacturers, who would be able to anticipate future demand. In this way, mechanisms such as the Global Drug Facility could create and support identifiable, transparent markets for internationally quality-assured drugs. Nonetheless, the Global Drug Facility's success in reducing prices was not universal: some second-line drugs, particularly kanamycin, cost substantially more from the Global Drug Facility than equivalent drugs of unknown quality offered on the private market. A key factor in the price of kanamycin was the limited availability of its active pharmaceutical ingredientonly a few suppliers met stringent World Health Organization quality criteria. Future interventions in the global dug market should address factors limiting the drug supply.
In addition, our analysis highlights the risks to any initiative based on consolidating demand such as the Global Drug Facility. For example, the facility's operations were affected by recent changes in funding. How might such risks be mitigated? First, the health of   the market for internationally qualityassured drugs depends on its size: a larger market can accommodate more manufacturers and promote competition as well as offering greater scope for economies of scale that will further reduce drug prices. It is, therefore, important to reverse the loss in the sales volume of first-line drugs we observed recently. Currently the Global Drug Facility supplies only the public sector (i.e. national tuberculosis programmes). However, the role of the private sector in controlling tuberculosis is being increasingly recognized and there may be new opportunities for the facility to supply internationally quality-assured drugs outside the public sector, where they are also needed. [19][20][21][22][23] Second, in addition to its current model of inviting applications for support from individual countries, the Global Drug Facility could also become a strong competitor if, in certain cases, it participated directly in national tenders (i.e. without a procurement agent) and became one supplier among many bidding to provide drugs for national tuberculosis programmes. If the Global Drug Facility received money from these programmes themselves, its reliance on donor support would be reduced. The large national tuberculosis programmes in India and South Africa could be important in this regard. On the donor side, our results highlight the risks of unstable funding sources and of funding coming from an increasingly small number of donors. However, it is important to note that donors have an influence that goes beyond their effect on purchasing power. For example, donor support encourages national tuberculosis programmes to adopt international guidelines (this is often a condition of support), ensures there is a pool of prequalified manufacturers who produce internationally quality-assured drugs and enables the Global Drug Facility to charge the lowest possible fees to participating countries, thus keeping costs low. Consequently, in the future, the Global Drug Facility should continue to serve public markets as it does at present, while at the same time seeking ways to relax constraints on the supply of tuberculosis drugs so that the facility can compete more directly in the tuberculosis drug market than it does at present. This combined approach could dramatically increase the level of demand managed by the Global Drug Facility, provide it with greater leverage and enable it to stimulate and sustain the market.
Our analysis has several limitations. The lack of fine-grained, countryspecific data from both IMS Health and the Global Drug Facility meant that we had to compare prices at the ex-works level rather than the patient level. Further, there may have been inaccuracies in IMS estimates of ex-works prices. However, if prices were underestimated, our finding that the Global Drug Facility negotiated prices that were lower than, or comparable to, those in the private market would be strengthened. However, if prices were overestimated, our   Only countries for which data were available on all drugs required for a full treatment course were included.  Our work suggests areas for future study. For example, apart from some research carried out in specific contexts, 10,20 few systematic, longitudinal studies have investigated the quality and quantity of drugs supplied outside the Global Drug Facility -that is, the funding channels represented by dashed lines in Fig. 1. There is, then, a need for systematic, large-scale surveys of the price, volume and quality of drugs available in different countries. This could be achieved by establishing national obser-vatories. Such data would be invaluable for building a comprehensive picture of the most cost-effective sources of tuberculosis drugs.
In conclusion, our analysis throws light on how the Global Drug Facility's operations since 2001 have influenced the dynamics of the market for internationally quality-assured tuberculosis drugs. Although challenging, it is essential that the global health community fully engages with such complex, global markets. The lessons learnt from the operation of the Global Drug Facility and other similar interventions will be invaluable in future discussions about the role of such models of engagement, to the common benefit of donors, governments and patients. ■

El Servicio Farmacéutico Mundial como una intervención en el mercado de los medicamentos contra la tuberculosis
Objetivo Investigar la financiación del Servicio Farmacéutico Mundial desde 2001 y analizar la influencia del Servicio en el precio de los medicamentos contra la tuberculosis de alta calidad. Métodos Los datos sobre el precio de los medicamentos contra la Resultados Por lo general, los medicamentos contra la tuberculosis con garantía de calidad suministrados por el Servicio Farmacéutico Mundial tenían un precio inferior que los medicamentos comprados en el sector privado. En 2012, tan solo tres fabricantes representaron 29,9 millones de dólares americanos (USD) de los 44,5 millones USD por el valor de los medicamentos de primera línea suministrados. El Fondo Mundial de Lucha contra el SIDA, la tuberculosis y la malaria proporcionó el 73 % (32,5 de los 44,5 millones de dólares americanos) y el 89 % (57,8 de 65,2 millones de dólares americanos) de los fondos para medicamentos de primera y segunda línea, respectivamente. Entre 2010 y 2012, la cuota de mercado del Servicio de medicamentos contra la tuberculosis de segunda línea aumentó del 26,1 % al 42,9 %, mientras que los precios disminuyeron en hasta un 24 % (de 1231 USD a 939 USD). Por el contrario, la cuota de mercado de medicamentos de primera línea del Servicio se redujo del 37,2 % al 19,2 % durante este tiempo, mientras que los precios aumentaron de 9,53 USD a 10,2 USD. Conclusión El precio de los medicamentos contra la tuberculosis suministrados a través del Servicio fue generalmente inferior que en el mercado privado. Sin embargo, para alcanzar su potencial y satisfacer las necesidades del mayor número de pacientes de tuberculosis, el Servicio requiere una financiación pública más diversa y estable, así como mayor flexibilidad para participar en el mercado privado.  a The relative price was the ratio of the mean global price of the drug on the private market to that of the drug supplied by the Global Drug Facility. Where this is 1, the two prices were equal.