Examining the trends in costs of medicines for drug-susceptible and drug-resistant tuberculosis from 2010 – 2013; an analysis of Global Fund PQR data
By: Ambrose Agweyu, Frank Wafula, Brian Mwangi, Cleopatra Mugyenyi, Kate Macintyreand Bernard Langat
14January2015
Table of Contents
Abbreviations
Preface
List of Tables
List of Figures
Executive Summary
Background
Methodology
Results
Costs of TB medicines (High burden countries vs other countries)
Analysis of trends
Discussion
Conclusion
References
Appendix 1: Box Plots Showing Extreme Values Excluded from Analyses of Trends
Appendix 2: Anti-TB Medicine Entries in PQR Database (January 2010 – December 2013)
Copyright © January 2015 by Aidspan. All rights reserved.
Abbreviations
FDCFixed Dose Combination
GDFGlobal Drug Facility
GFGlobal Fund
HBCHigh burden countries
LBCLow burden countries
NTPNational Tuberculosis Program
PASPara-Aminosalicylate
PQRPrice and Quality Reporting
RHZ Rifampicin, Isoniazid, Pyrazinamide
RHZERifampicin, Isoniazid, Pyrazinamide, Ethambutol
TBTuberculosis
WHOWorld Health Organization
Preface
Aidspan ( is an international non profit organization based in Kenya, whose mission is to reinforce the effectiveness of the Global Fund. Aidspan performs this mission by serving as an independent observer of the Fund and by providing services that can benefit all countries wishing to obtain and make effective use of Global Fund financing.
This paper is one of several Aidspan research reports available at Reports published over the past year by Aidspan include:
- The New Funding Model Allocations: An Aidspan Analysis (November 2014)
- Expenditure reported by national Tuberculosis programs in 22 high-burden countries between 2010 – 2012: what is the Global Fund’s contribution? (Oct 2014)
- Options for reforming the Global Fund Board (April 2014)
- Conflict of Interest in Country Coordination Mechanisms: An Aidspan Survey (May 2014)
- Procurement Cost Trends for Global Fund Commodities: Analysis of Trends for Selected Commodities 2005–2012 (April 2013)
- Global Fund Principal Recipient Survey: An Assessment of Opinions and Experiences of Principal Recipients (April 2013)
Aidspan also publishes news, analysis and commentary articles about the Global Fund in its Global Fund Observer (GFO)newsletter and on GFO Live. To receive GFO Newsletter, send an email to .
Aidspan finances its work primarily through grants from governments and foundations. Aidspan does not accept funding of any kind from the Global Fund. Aidspan and the Global Fund maintain a positive working relationship, but have no formal connection. Aidspan does not allow its strategic, programmatic or editorial decision-making to be influenced by the Global Fund or by relationships with Aidspan’s funders.
Acknowledgements
Aidspan thanks the following donors whoall contributed to supporting Aidspan in 2014, and thus supporting the production of this paper: the UK Department for International Development (DFID), the Ford Foundation, GIZ Backup Initiative, Irish Aid, the Government of the Netherlands, NORAD and Hivos.
List of Tables
Table 1: Commodities included in the analysis
Table 2: Comparison of Median Unit Costs of Selected Anti-TB Medications Purchased by High and Low Burden Countries
List of Figures
Figure 1: Trends in Costs of Adult RHZE FDC from 2010 to 2013 (High versus Low Burden Countries)
Figure 2: Trends in Costs of Adult TB Cat I+III Patient Kit A from 2010 to 2013 (High versus Low Burden Countries)
Figure 3: Trends in Costs of Pediatric RHZ from 2010 to 2013 (High versus Low Burden Countries)
Figure 4: Trends in Costs of Adult RHZE versus Pediatric RHZ from 2010 to 2013
Figure 5: Trends in Costs of Capreomycin from 2010 to 2013 (High versus Low Burden Countries)
Figure 6: Trends in Costs of Cycloserine from 2010 to 2013 (High versus Low Burden Countries)
Figure 7: Trends in Costs of Levofloxacin from 2010 to 2013 (High versus Low Burden Countries)
Figure 8: Trends in Costs of PAS Sodium from 2010 to 2013 (High versus Low Burden Countries)
______
An analysis of Global Fund PQR data
An Aidspan Publication
14 Jan 2015 Page 1
Executive Summary
Tuberculosis (TB) is a major public health challenge affecting up to one third of the world’s population. Despite the availability of effective treatments, TB continues to claim 1.5 million lives annually. Substantial reductions in incidence and mortality can be achieved through improved access to effective anti-TB medications. This strategy is heavily supported by the Global Fund against AIDS, Tuberculosis and Malaria (GF) through grants disbursed to national TB programs, which are estimated to support up to three quarters of funding for programs in high burden countries. This report describes the trends in costs for selected TB medications purchased through grants from the GF from 2010 – 2013 in the 22 high burden countries (HBC) and low burden countries (LBC).
The median cost of all first line anti-TB medications was significantly lower in HBC compared to LBC with pediatric formulations costing approximately half the value of adult alternatives. In contrast, the costs of second line treatments showed no variation across high and low burden countries. Further, the unit costs of second line treatments were up to 100 times higher than those of first line formulations. A rising trend was observed for first line treatments over the four year period while the costs of second line treatments, with the exception of capreomycin, reduced over time.
Trends in the costs of second line treatments and first line treatments in low burden countries may be attributed to the role of the Global Drug Facility (GDF) in obtaining competitive prices through pooled procurement. Conversely, the costs of first line treatments in high burden countries show characteristics of a mature market. Differences in costs for pediatric and adult first line medications coincide with the disbursement of grants by UNITAID to subsidize the cost of pediatric anti-TB drugs. This analysis illustrates important differences in costs of anti-TB medications and highlights potential areas of intervention for initiatives aimed at accelerating progress towards achieving global TB targets through increasing access to affordable anti-TB medications.
Background
Tuberculosis continues to rank among the leading causes of death accounting for an estimated 9 million new annual infections and 1.5 million deaths globally (1). Along with HIV and malaria, tuberculosis was acknowledged as a global public health challenge by world leaders in the Millennium Declaration and articulated under Millennium Development Goal 6 (2). Efforts towards achieving this goal have been rewarded with variable degrees of success with 8 out of 22 of the high burden countries (HBC)[1] having met all of the 2015 targets for reductions in TB cases and deaths and a further two on track to do so (1). However, the average global decline in TB incidence between 2000 and 2013 remains modest at 1.5% per year.
Increasing access to effective anti-TB medicines is a major strategy in the effort to reduce the burden of TB. Although financing for anti-TB medicines from both domestic sources has increased steadily over recent years, donor funding remains a major source of funding for anti-TB medicines, accounting for up to 70% of the national tuberculosis program (NTP) budgets in some high burden countries (3). In a recent report describing the expenditure of national TB programs (NTPs) of the 22 HBC between 2010 – 2012, the Global Fund against AIDS, Tuberculosis and Malaria (GFATM) was reported to have contributed up to 76% of total national expenditure of the programs of the countries studied (4). Despite its central role in financing TB care, literature on the trends in costs for anti-TB medicines purchased through the Global Fund is surprisingly scarce.
Methodology
We sought to describe the trends in procurement cost for selected TB commodities using information collected through the Global Fund’s Price Quality Reporting (PQR) system from 2010-2013. The PQR is a Global Fund provided online reporting facility that tracks commodity prices as reported by countries as they buy commodities using their grant money. We specifically sought to analyze the trends of first line and second line TB medicines procured by both the 22 high burden and all other countries that received grants from the Global Fund between 2010 and 2013.
Data from the PQR were obtained from the Global Fund, and inspected for among other things, completeness, presence of outliers and distribution. Extreme values (after inspection) were assumed to be data entry errors, and were excluded from the analyses of trends. Details of the outliers excluded are provided in Appendix 1.The scope of the analysis was defined a priori by restricting the comparisons to those of the costs reported for first line anti-TB fixed dose combinations (adult and pediatric) and frequently purchased single medications for drug resistant TB in high burden and low burden countries (table 1). The complete list of anti-TB medications contained in the database is provided in Appendix 2.
Table 1: Commodities included in the analysis
Commodity / Number of entries in PQR databaseHigh burden countries / Other countries / Total
First line
Adult RHZE1 FDC2 / 82 / 143 / 225
TB Cat. I+III Patient Kit A3 / 31 / 22 / 53
Pediatric RHZ4 FDC / 22 / 40 / 62
Second line
Cycloserine / 103 / 285 / 388
Levofloxacin / 113 / 277 / 390
Capreomycin / 53 / 165 / 218
PAS Sodium / 66 / 212 / 278
1 RHZE - Rifampicin 150mg + Isoniazid 75mg + Pyrazinamide 400mg + Ethambutol 275mg
2FDC – Fixed Dose Combination
3 Four-drug FDC 6x28tabs(R 150mg / H 75mg / Z 400mg / E 275mg) plus Two-drug FDC 12x28tabs (R 150mg / H 75mg)
4 Pediatric RHZ - Rifampicin 60mg + Isoniazid 30mg+ Pyrazinamide 150mg
Unit costs were derived from the quantities of medications purchased and total costs paid for respective items purchased. Median costs for the selected medicines procured from 2010 to 2013 were calculated with interquartile ranges. Comparisons between high burden countries and other countries were made using the Mann-Whitney U test and corresponding p-values reported. Trends over the period of interest were illustrated using scatter plots. Unit costs were regressed against time and represented as lines of best fit superimposed on the scatter plots. We conducted hypothesis tests to compare the regression coefficients of costs for the selected commodities in high and low burden countries. All analyses were conducted using STATA version 13 (Stata Corp, Texas, USA).
Results
At the time of analysis, the PQR database contained 27,748 entries from 128 countries. Of these, 5,280 (19.0%) were categorized as TB-related medications. All high burden countries were represented, with the exception of South Africa and Brazil, both of whom did not procure TB medicines through the Fund’s facility during the period covered in the analysis.
Costs of TB medicines (High burden countries vs other countries)
Table 2 summarizes a comparison of costs for selected first and second line TB medicines procured through the Global Fund from between 2010 and December 2013.
Median costs for RHZE, the first line adult fixed dose combination (FDC) showed minimal variation ranging from USD .05 to .06 per tablet. A comparison of overall costs between high and low burden countries revealed significantly lower median costs in the former category (P.001). Similar differences were observed for TB Category I+III Patient Kit A (a kit combining first line medicines for the intensive and continuation phases of treatment - see table 1). The median cost for the other first line treatment analyzed - pediatric RHZ FDC was also lower within the high burden countries, although the association was relatively weaker (P=.07).
In contrast, there was no relationship between costs and disease burden for the four second line treatments included in this analysis. With the exception of levofloxacin (a broad spectrum antibiotic), costs of second line treatments were substantially higher than first line treatments. The cost of Capreomycin, a second line anti-TB, was nearly 100 times higher than that for the adult four-drug FDC.
Notable variations were observed in the costs reported for pediatric RHZ FDC and capreomycin (IQR .03 to .07 and 4.3 to 8.0 respectively). In comparison, the costs of treatments within high burden countries showed only minimal variation, reflected by narrow interquartile ranges of median unit costs.
______
An analysis of Global Fund PQR data
An Aidspan Publication
14 Jan 2015 Page 1
Table 2: Comparison of Median Unit Costs of Selected Anti-TB Medications Purchased by High and Low Burden Countries
Category / First line medications / Second line medicationsAdult RHZE FDC / TB Cat. I+III Patient Kit A / Pediatric RHZ FDC / Capreomycin / Cycloserine / Levofloxacin / PAS Sodium
High burden countries / .06 (83)*
(.05 – .06)** / 18.2 (31)
(17.1 – 20.2) / .03 (22)
(.03 – .03) / 5.5 (53)
(4.3 – 5.7) / .59 (105)
(.59 – .59 ) / .06(117)
(.05– .07) / .39 (79)
(.38 – .43)
Other countries / .06 (149)
(.06 – .06) / 20.2 (22)
(19.1 – 20.2) / .04 (41)
(.03 – .07) / 5.5 (165)
(4.3 – 8.0) / .59 (293)
(.59 – .59) / .06 (278)
(.05– .06) / .38 (253)
(.38 – .41)
P value† / .001 / .05 / .07 / .55 / .90 / .67 / .30
* Total number of entries ** Interquartile range
†Mann-Whitney U test
______
An analysis of Global Fund PQR data
An Aidspan Publication
14 Jan 2015 Page 1
Analysis of trends
First line Adult RHZE FDC
Significant ascending linear trends were observed for the cost of first line adult RHZE FDC from 2010 to 2013 in both HBC (P<.001) and other countries (P<.001). The change in price over time did not vary between the two groups (P=.53) (figure 1).
Figure 1: Trends in Costs of Adult RHZE FDC from 2010 to 2013 (High versus Low Burden Countries)
First line TB Cat. I+II Patient Kit A
Although the number of entries for the adult TB Category I+II Patient Kit A was considerably lower than that recorded for the RHZE FDC, a similar trend of progressively rising costs was observed for purchases in both HBC and other countries (P< .001 in both groups) (figure 2). The change in price over time also did not vary between the two groups for this commodity.
Figure 2: Trends in Costs of Adult TB Cat I+III Patient Kit A from 2010 to 2013 (High versus Low Burden Countries)
First line Pediatric RHZ FDC
The costs of the pediatric RHZ FDC increased over time in both high and low burden countries. This trend, although statistically significant (P=.03 and P=.02 for high and low burden countries respectively), was less pronounced than that observed for costs of other first line medications (figures 2 and 3). The change in costs over time in HBC versus other countries was almost identical (P=.91) (figure 3).
Figure 3: Trends in Costs of Pediatric RHZ from 2010 to 2013 (High versus Low Burden Countries)
Pediatric versus Adult First line FDCs
A comparison between pediatric and adult first line FDCs similarly revealed a rising trend in costs over time. Although the average costs of both the adult medications was noticeably higher than the pediatric alternative, the increase in price over time was very similar between the two groups (P=.54) (figure 4).
Figure 4: Trends in Costs of Adult RHZE versus Pediatric RHZ from 2010 to 2013
Second line medications: Capreomycin
Costs of capreomycin increased in both high and low burden countries over time. Although statistically significant upward linear trend was observed (P=.01 and <.001 for high and low burden countries respectively), a sudden rise in the cost of individual purchases was apparent in 2011 with a reduction in subsequent years (figure 5).
Figure 5: Trends in Costs of Capreomycin from 2010 to 2013 (High versus Low Burden Countries)
Second line medications: Cycloserine
The cost of cycloserine reduced steadily over the four year period in both high and low burden countries (P<.001). Although the direction of change in cost was consistent in the two groups, the rate of decline appeared to be more pronounced in low burden countries (P=.05) (figure 6).
Figure 6: Trends in Costs of Cycloserine from 2010 to 2013 (High versus Low Burden Countries)
Second line medications: Levofloxacin
A significant decline in the cost of levofloxacin was observed within HBC over the four years (P=.002). Within low burden countries the cost of levofloxacin remained unchanged over time (P=.36). This difference in trend between the two groups of countries was statistically significant (P=.02) (figure 7).
Figure 7: Trends in Costs of Levofloxacin from 2010 to 2013 (High versus Low Burden Countries)
Second line medications: PAS Sodium
Consistent with trends in costs of other second line anti-TB medications, the costs of PAS Sodium also showed substantial reductions over time (P<0.01 in both the high burden and the other countries). A comparison of the fitted trends revealed that the difference in the rate of decline did not vary significantly between the two categories of countries (P=.18) (figure 8).
Figure 8: Trends in Costs of PAS Sodium from 2010 to 2013 (High versus Low Burden Countries)
Discussion
We used price information reported through the Global Fund’s PQR system to study trends in costs of TB medicines between January 2010 and December 2013. We believe this is the first time such an analysis has been done. Previous analyses have looked at the trends for HIV and malaria commodities (5, 6).
The Global Drug Facility (GDF), a section under the WHO-supported STOP TB initiative (7) has contributed significantly towards ensuring access to high quality affordable anti-TB medicines since its inception in 2001 (8). This facility, which functions through pooled procurement of competitively sourced commodities, aims to ensure uniform costs across countries for supported commodities which include first and second line anti-TB medications (including pediatric drugs). This analysis reveals important differences between costs across commodities purchased in the 22 high burden countries and low burden countries. Our findings suggest the existence of external channels of procurement not supported by the GDF. Indeed a recently published report estimated that the GDF supplied first line treatments for only a third of reported TB cases in 2011 (9). This same report states that the market share of GDF-supplied first line treatments has been shrinking over time while that for second line treatments appears to be growing. This trend may be attributed to a vibrant market for first line anti-TB medications purchased directly from suppliers at competitive costs – a situation that supports the continued need for rigorous quality assurance mechanisms such as the WHO prequalification system for suppliers. The GDF however maintains an important role in the procurement of second line TB medicines with a limited number of manufacturers and for first line medicines in low burden countries.
The treatment of drug resistant tuberculosis is a growing global challenge. This analysis revealed an encouraging general downward trend in the costs of medications for drug resistant tuberculosis. This finding is consistent with the 2013 annual report of the Stop TB Partnership which attributes the declining cost to the GDF (10).In spite of the declining costs of second line treatments, the cost of treating multi-drug resistant TB remains over 100 times that of treating drug sensitive tuberculosis (11). A significant proportion of this amount is spent on injectable medications. The analysis showed, for instance, that the cost of capreomycin increased over the period of analysis. This rise was attributed to a global shortage in 2011 that followed an acute shortage of the active pharmaceutical ingredient from the sole quality assured manufacturer (12). High costs of this and other second line drug illustrate the immediate need for expansion of the pool of quality assured manufacturers for second line treatmentswhich currently consumes a disproportionate amount of national budgets for TB and threatens to reverse the gains made in lowering the burden of drug susceptible TB over recent years. The relatively low cost of levofloxacin, a second line fluoroquinolone, is reflective of its availability from at least three suppliers (13), and use beyond TB treatment as a broad spectrum antibiotic with several other clinical indications.