Financing HIV/AIDS programs in sub-Saharan Africa

Health Affairs

Volumn 28, Issue 6, 2009, Pages 1606-1616

  Haacker, Markus ^a  

^a LONDON SCHOOL OF HYGIENE AND TROPICAL MEDICINE   (United Kingdom)

Author keywords

[No Author keywords available]

Indexed keywords

ACQUIRED IMMUNE DEFICIENCY SYNDROME; AFRICA; CONFERENCE PAPER; ECONOMIC ASPECT; HEALTH CARE COST; HEALTH CARE FINANCING; HEALTH PROGRAM; HUMAN; HUMAN IMMUNODEFICIENCY VIRUS INFECTION; LOWEST INCOME GROUP; NONHUMAN; PREVALENCE;

ACQUIRED IMMUNODEFICIENCY SYNDROME; AFRICA SOUTH OF THE SAHARA; HEALTH CARE COSTS; HEALTH EXPENDITURES; HIV INFECTIONS; HUMANS; PREVALENCE;

EID: 77951091206     PISSN: 02782715     EISSN: 15445208     Source Type: Journal    
DOI: 10.1377/hlthaff.28.6.1606     Document Type: Conference Paper

References (28)

1. World Bank. Country classification [Internet]. Washington: World Bank; 2009 [cited 2009 Sep 23]. Available from: http://go.worldbank.org/K2CKM78CC0
2. In South Africa, important cost categories (medical expenditures not related to antiretroviral treatment, certain social expenditures) are not included in the estimates of HIV/AIDS-related spending.
3. UNAIDS. 2008 report on the global AIDS epidemic. Geneva: UNAIDS; 2008.
4. Organization for Economic Cooperation and Development. Creditor reporting system, aid activity database. Paris: OECD; 2009.
5. This analysis was based on data on external financing from UNGASS reports (Note 3). The data in Note 4 are an alternative source of data on external financing, available for a larger number of countries. This paper focuses on the UNGASS data to ensure consistency with spending data.
6. The exception is Gabon, where external financing accounts for about 40 percent of HIV/AIDS-related spending. However, this spending in Gabon is less than US$10 million, or 0.1 percent of GDP.
7. World Health Organization. WHO statistical information system (WHOSIS). Geneva: WHO; 2009 [dted 2009 Aug 29]. Available from http://www.who.int/whosis/ en/index.html
8. (Unweighted) average related to thirty-two African low-income countries for 2006, based on Note 7.
9. Regression analysis to complement the discussion is available in an online appendix at http://content.healthaffairs.org/cgi/content/full/28/1606/ DCI.
10. For example, the Global Fund used to demand a certain proportion of domestic financing of proposals, depending on the income level and the level of HIV prevalence in the applicant country.
11. Equation (1) in the online appendix shows this evidence; see Note 9.
12. UNAIDS. Annual estimates of HIV prevalence, people living with HIV/AIDS, and HIV/AIDS-related deaths, 1990-2007. Geneva: UNAIDS; 2008.
13. United Nations Population Division. World population prospects: The 2008 revision - population database. New York United Nations; 2009.
14. A visual depiction of this model is available in the online appendix, as in Note 9.
15. Stover J. Spectrum - estimating the need for ART and the resources required Paper presented at Department of Health ART Costing Workshop; 2009 Mar; Pretoria, South Africa.
16. The average transition time from infection to first-line treatment need was set at 7.7 years. Without treatment, death occurs 3.5 years after the onset of treatment need (these numbers were set to match UNAIDS summary data). The average time from initiation of first-line treatment to need for second-line treatment was set at eight years, and the average duration of second-line therapy was set at 10.5 years. One complication that this modd cannot address is that the approach implies that treatment commences at the time of need, whereas the times may differ individually and on an aggregate level, especially as the criteria applied by UNAIDS for estimating treatment need could be different from national guidelines.
17. For people living with HIV/AIDS but not requiring treatment, this is obtained as an average over survival curves for different ages, based on the assumption that (most) HIV infections occur between ages twenty and thirty-five. For the groups receiving first- and second-line therapy, the survival curves were shifted using average transition times.
18. McCarthy O, Over M. Projecting the future budgetary cost of AIDS treatment in poor countries: a manual for the AIDS cost computer programs. Washington (DC): Center for Global Development; 2009.
19. Over M. Prevention failure: the ballooning entitlement burden of U.S. global AIDS treatment spending and what to do about it. Washington (DC): Center for Global Development; 2008. Working paper no. 144.
20. For these projections, the framework was modified in two directions. First, for the sample of countries, the upper limit of the costs of second-line therapy used by McCarthy and Over (Note 18) - just below US$5,000) seems too high, Therefore, the projection used an upper limit of US$2,200. Second, prices of second-line drugs are expected to fall; the projections assumed that they will decline through 2014 at the rate implied by Stover (Note 15).
21. This assumption is broadly consistent with thorough costing estimates available for some countries, as well as the literature on the economic impacts of HIV/AIDS preceding the scaling up of antiretroviral therapy, usually assuming costs of treatment amounting to one to four times GDP per capita. (The parameter is lower here; this paper focuses on fiscal costs, excluding private spending.) Because HIV/AIDS mortality does not fluctuate widely, the projections are robust with respect to the distribution of the individual costs of treatment (the analysis assumes two years before death) over time.
22. Data on the composition of spending, by major category, are available from UNAIDS (see Note 3) for only twenty-one of the thirty-four countries in Exhibit 1.
23. These countries are Angola, Benin, Botswana, Burkina Faso, Burundi, Cameroon, Central African Republic, Chad Congo, Côte d'Ivoire, Eritrea, Ethiopia, Gabon, Ghana, Guinea, Kenya, Lesotho, Liberia, Malawi, Mali, Mozambique, Namibia, Niger, Nigeria, Rwanda, Senegal, Sierra Leone, South Africa, Swaziland, Tanzania, Togo, Uganda, Zambia, Zimbabwe. The most common reason for exclusion from the data set is the fact that UNAIDS (see Note 12) truncates or does not report small absolute values, so that sufficient data on HIV prevalence and HIV/AIDS-related deaths are not available.
24. To avoid double-counting HIV/AIDS-related expenditures in the numerator and denominator, the levd of total and public health expenditures for 2000 (as reported by the World Health Organization; see Note 7) was used as a basis for comparison.
25. Because drugs are imported it makes sense to exclude them (as was done for many other imported supplies for which data are not adequate) from the comparison to obtain an indicator for the scaling up of health services implied by the projections.
26. This is specified in Equation (2) in the online appendix, as in Note 9.
27. The level of HIV prevalence explains 65 percent of the variation in total spending and 75 percent of the variation in domestically financed spending in 2015.
28. Further discussion is available from Haacker M. The impact of HIV/AIDS on government finance and public services. In: Haacker M, editor. The macroeconomics of HIV/AIDS. Washington (DC): International Monetary Fund 2004. p. 41-95.