Abstract
Achieving the 90-90-90 targets by 2020 requires increased focus, resources, and efficiency to provide earlier access to antiretroviral therapy (ART).
Methods:
We used 2009 to 2013 National AIDS Spending Assessment data to assess HIV care and treatment spending in 38 high-burden, low- and middle-income countries (LMICs).
Results:
In 2013, 23 of the 38 high-burden countries spent less than 50% of total HIV spending on care and treatment. HIV spending on ART per people living with HIV (PLHIV; adjusted) averaged US$299 (US$32-US$2463). During 2009 to 2013, a 10% increase in average spending on care and treatment per PLHIV was associated with an increase in ART coverage of 2.4% and a decrease in estimated AIDS-related death rate of 2.4 per 1000 PLHIV.
Discussion:
HIV spending in high-burden LMICs does not consistently reflect the new science around the preventative and clinical benefits of earlier HIV diagnosis and ART initiation.
Introduction
Over 30 years after the first case was reported, HIV remains a significant public health problem. More than 39 million people have died of AIDS and there are an estimated 36.9 million (34.3 million-41.4 million) people living with HIV (PLHIV). 1 There were around 2 million (1.9 million-2.2 million) new HIV infections and 1.2 million (980 000-1.6 million) AIDS-related deaths in 2014. 1 Scientific advances have yielded treatment and prevention interventions that can help achieve the goal of controlling the HIV pandemic. Antiretroviral therapy (ART) is a powerful intervention that prevents disease progression, opportunistic infections (OIs), tuberculosis (TB), mortality, and HIV transmission. 2 -10 Studies suggest that treatment should be started irrespective of CD4 count and that expanded access to treatment in some settings could lead from control of HIV to the eventual elimination of HIV. 4,5,10,11 From 1995 to 2013, ART averted an estimated 7.6 million AIDS-related deaths globally. 12 Despite the progress in expanding ART to 15 million people by the end of March 2015, only 41% (38%-46%) of the estimated 37 million PLHIV were receiving ART. 1
International and domestic funding for the HIV/AIDS response has been significant. From 2004 onward, the Global Fund to Fight AIDS, TB, and Malaria has disbursed US$73.2 billion for the AIDS response, and the President’s Emergency Plan for AIDS Relief (PEPFAR) has invested US$46.6 billion in bilateral HIV/AIDS programming. 13,14 In 2014, an estimated US$20.2 billion was invested in the AIDS response in low- and middle-income countries (LMICs). 1 When international and domestic funding is combined, some high-burden LMICs’ spending exceeds half a billion dollars a year or between US$175 and US$3100 per PLHIV. This unprecedented public health investment raises legitimate concerns regarding prioritization, accountability, and impact. 15
In 2014, the Joint United Nations Programme on HIV/AIDS (UNAIDS) called for 90-90-90 targets for 2020: 90% of PLHIV aware of their HIV status, 90% of them on sustained ART, and 90% of those on ART will have viral suppression. 16,17 Antiretroviral therapy prevents illness, death, transmission, and costs, and reaching the 90-90-90 targets will have a significant impact on the HIV epidemic. 17 However, countries are facing significant challenges in identifying the maximum number of PLHIV through testing interventions and subsequently engaging them in care and treatment services. International and national policies around HIV diagnosis and access to ART vary considerably, 18,19 though there is a growing call for harmonizing such policies, which may reorient the existing HIV service delivery models toward earlier diagnosis and ART initiation test and treat. This pivot to the 90-90-90 targets and subsequent result—more PLHIV diagnosed and placed onto ART—has financial implications, in terms of both expense and savings. While requiring a reprioritization and an increase in spending, the move toward earlier diagnosis and treatment will save money over time due to its prevention effects. The savings fall into 3 general categories: (1) health system savings (eg, healthier clients and community-based service delivery including fewer clinic visits, personnel, task sharing, reduced need for laboratories; fewer OIs, cancers, and hospitalizations; and fewer newly infected people living requiring services), (2) savings for PLHIV (eg, reduced time and transport costs due to fewer clinic and pharmacy visits; fewer payments for services, commodities, and medications; increased productivity and income), and (3) society savings (eg, healthier community, return on education and development investment in individuals through continued employment, contribution to growth, development and the economy, intact families and social networks, and payment of taxes). Therefore, how countries choose to allocate and prioritize their HIV funding will impact the scale-up of HIV testing and treatment services for PLHIV in the future and have a direct impact on whether the HIV epidemic is controlled and ultimately eliminated. This article examines the pattern, source, determinants, and impact of HIV spending on care and treatment from 2009 to 2013 in 38 LMICs with the highest burden of HIV in 2013.
Methodology
Data Collection
The UNAIDS maintains a database of country-reported data—AIDSinfo—which includes HIV spending levels based on National AIDS Spending Assessment (NASA) classifications. 20 UNAIDS collects country NASA reports every other year for its United Nations General Assembly Special Session (UNGASS) reporting. The NASAs are HIV focused and designed to describe the financial flows and expenditures using the same categories as the globally estimated resource needs. This alignment was conducted in order to provide necessary information on the financial gap between resources available and resources needed and to promote the harmonization of different policy tools that are frequently used in the AIDS field. NASA provides indicators of the financial country response to AIDS and supports the monitoring of resource mobilization. Thus, NASA is a tool to install a continuous financial information system within the national monitoring and evaluation framework. NASA is not an accounting system insofar as it tracks spending as reported by countries. Donor and government spending are divided in NASA into 8 spending classes or chapters of AIDS spending categories: prevention, care and treatment, orphans and vulnerable children, strengthening program management and administration, incentives for human resources, social protection and social services, enablement of environment and community programs, and research.
From 2009 to 2013, we abstracted the available data for LMICs with the highest burden of HIV in 2013 and greater than 75 000 PLHIV (41 countries in total, representing an estimated 84% of global HIV burden and 90% of HIV burden in LMICs in 2013). 12 India, Lesotho, and the United Republic of Tanzania were excluded due to the lack of availability of complete data. For the remaining 38 countries (73% of the global burden), which were all LMICs, we analyzed and compared the detailed data on HIV spending from both domestic and international sources for the following program areas: (a) care and treatment (spending on ART is included in the care and treatment category), (b) prevention, (c) program management and administration strengthening, and (d) others (spending on orphans and vulnerable children, incentives for human resources, social protection and social services, enabling environment, and research). Expenditure on care and treatment includes the costs of HIV testing, prophylaxis and treatment for OIs, ART, nutrition support, HIV-related laboratory monitoring, psychological treatment, support services, dental services, transportation, home-based care, informal care, and traditional medicines. The ART expenditure includes the cost of World Health Organization (WHO)-recommended antiretroviral drugs, supply logistics, and the entire spectrum of ART service delivery (including the cost of human resources) for both adults and children. 21 Estimates on HIV prevalence, new HIV infections, PLHIV receiving ART, and AIDS-related deaths were also obtained from AIDSinfo. 20 AIDS-related deaths are calculated using a variety of parameters including ART coverage and the estimated death rates on and off treatment. Data on gross domestic product (GDP) and per capita income came from the World Bank database. 22
We calculated 2 additional indicators—“HIV spending on care and treatment per PLHIV” and “HIV spending on ART per PLHIV.” Both of these indicators were adjusted using the national price level (price level ratio of purchasing power parity conversion factor to market exchange rate) 22 to correct for the differences in prices of goods and services in countries. The ratio tells how many dollars are needed to buy a dollar worth of goods in the country as compared to the United States. We used correlation scatter plot and univariate regression analysis (in Microsoft Excel 2013) to look at per capita income and HIV prevalence as predictors of HIV spending on care and treatment per PLHIV. 23 We also calculated “cost of treatment per person on ART” defined as total HIV spending on ART divided by people receiving ART.
Impact of “average HIV spending on care and treatment per PLHIV” (between 2009 and 2013) on ART coverage and AIDS-related death rate in 2013 was also examined using correlation scatter plot and univariate regression analysis (in Microsoft Excel 2013). The 2014 data for ART coverage and AIDS-related deaths were available later but for fewer LMICs; hence, we restricted the analysis to 2013 data. The average HIV spending on care and treatment per PLHIV was calculated over a varying number of years depending on the availability of data per country and adjusted using national price levels. The ART coverage was calculated as PLHIV reported to be on ART at the end of the year divided by the estimated number of PLHIV in the same year. This calculation provides a more comparable estimate of coverage across years since the WHO HIV guidelines’ ART eligibility criteria were amended from CD4 count <200 cells/mm3 in 2006 to CD4 count <350 cells/mm3 in 2009, CD4 count <500 cells/mm3 in 2013, and irrespective of CD4 count in 2015 (or “test and treat”). 24 -27 The AIDS-related death rate was defined as the number of AIDS-related deaths per 1000 PLHIV.
Finally, we looked at trends in HIV spending on care and treatment, prevention, program management and administration strengthening, and other categories over the period 2009 to 2012. Data were available from 33 countries for 2009 (South Africa had incomplete data and was excluded for this trend analysis), 29 countries for 2010, 33 countries for 2011, and 22 countries for 2012. We also looked at trends in HIV spending by program areas, AIDS-related deaths and ART coverage for 2 countries with diverse epidemics—Nigeria (the country with the second highest burden of HIV with 9% of PLHIV in 2013) and Swaziland (the country with the highest HIV prevalence of 28% in 2013).
Results
Total HIV Spending and Spending on HIV Care and Treatment by Country
The latest available data on total HIV spending and spending on HIV care and treatment for the 38 countries is shown in Table 1. South Africa, the country with the highest HIV burden (18% of estimated PLHIV), had a total HIV spending of US$2.3 billion (2009 data), with US$1.6 (70%) billion spending on care and treatment. Comparatively, total HIV spending in Nigeria (9% of PLHIV) was US$570 million, with US$190 (33%) million for care and treatment. HIV spending as a percentage of GDP varied significantly between 0.01% in Indonesia to 2.8% in Haiti. Of the 38 countries, the percentage of total HIV spending allocated for care and treatment ranged from 10% to 75%. Of the 38 countries, 23 countries (32% of global burden) spent less than 50% of their HIV spending on care and treatment (Figure 1). Similarly, the proportion of HIV spending on ART was below 50% in 31 countries, with 5 countries below 15%.
HIV Spending on Care and Treatment Including ART for 38 Low- and Middle-Income Countries with the Highest Burden of HIV in 2013.
Abbreviations: ART, antiretroviral therapy; PLHIV, person living with HIV; UNAIDS, the Joint United Nations Programme on HIV/AIDS; NA, not applicable.
aData obtained from UNAIDS Aidsinfo data tool.
bHIV spending on care and treatment (or ART) per PLHIV were divided by national price levels to adjust for price differences between countries.
Proportion of HIV spending on care and treatment and HIV spending on care and treatment per PLHIV (adjusted) in 38 low- and middle-income countries, 2009 to 2013. We have used the latest available data on HIV spending for the countries. PLHIV indicates people living with HIV.
Source of HIV Spending on Care and Treatment: Domestic and International Sources
According to the latest data reported from 38 countries, an average of 68% of funding for HIV care and treatment originated from domestic sources. However, 19 countries relied on international funding for more than 75% of their spending on care and treatment, with 5 countries funding their care and treatment spending 100% from international sources (Table 1). On an average, lower-income countries had a higher proportion of HIV care and treatment spending financed from international sources (correlation coefficient = −0.8).
HIV Spending on Care and Treatment and on ART per PLHIV, and Cost of Treatment per Person on ART
HIV spending on care and treatment and ART per PLHIV was highest in Mexico at US$2056 and US$1527, respectively (Table 1). The HIV spending on care and treatment per PLHIV across all countries whose spending was analyzed was US$270 on average, ranging from US$17 in South Sudan to US$2056 in Mexico. The average of HIV spending on ART per PLHIV was $152 (range US$16-US$1527). Cost of treatment for people on ART varied from US$110 in Zimbabwe to US$3619 in Mexico. After accounting for price differences in countries, HIV spending on care and treatment per PLHIV (adjusted) varied from US$33 in South Sudan to US$3316 in Mexico. In most of the countries with HIV prevalence between 1% and 10%, HIV spending on care and treatment per PLHIV (adjusted) was below US$200 (Figure 1). HIV spending on ART per PLHIV (adjusted) averaged US$299, with a range of US$32 in South Sudan to US$2463 in Mexico.
Per Capita Income, HIV Prevalence, and HIV Spending on Care and Treatment per PLHIV
HIV spending on care and treatment per PLHIV (adjusted) was found to be positively but weakly correlated to per capita income (Figure 2). The degree of dispersion in care and treatment spending per PLHIV (adjusted) was high among the LMICs. While all the upper-middle-income countries were spending higher amounts on care and treatment per PLHIV, some lower-income countries (Burundi, Haiti, Ethiopia, Rwanda, and Uganda) were spending similar amounts on care and treatment per PLHIV. In these countries, more than 75% of the funding for care and treatment came from international sources. HIV spending on care and treatment per PLHIV (adjusted) was extremely low in Angola, Cameroon, Chad, Myanmar, and South Sudan. Univariate regression analysis showed that the effect of per capita income on HIV spending on care and treatment per PLHIV (adjusted) was statistically significant (P = .009). There was no correlation between HIV prevalence and HIV spending on care and treatment per PLHIV (figure not shown).
Scatter plot showing correlation between per capita income and HIV spending on care and treatment per person living with HIV (adjusted). We have used the latest available data on HIV spending for the countries. Values in parentheses show the P values.
Association of Average HIV Spending on Care and Treatment per PLHIV (2009-2013), with ART Coverage and AIDS Death Rate in 2013
There was a positive correlation between average spending on care and treatment per PLHIV between 2009 and 2013 (adjusted) and ART coverage in 2013 (Figure 3). The AIDS death rate in 2013, estimated using ART coverage figures, was negatively correlated with average spending on care and treatment per PLHIV (adjusted; Figure 4). A 10% increase in average spending on care and treatment per PLHIV (adjusted) during this period was associated with an increase in ART coverage by 2.4 percentage points and a decrease in AIDS-related deaths by 2.4 for every 1000 PLHIV. Univariate regression analyses showed that the effect of average HIV spending on care and treatment per PLHIV (2009-2013 and adjusted) on both ART coverage (2013) and AIDS death rate (2013) was statistically significant (P <.0001).
Correlation between average spending on care and treatment per PLHIV—adjusted (between 2009 and 2013) and ART coverage in 2013. Average spending on care and treatment for countries are calculated across a varying number of years depending on the availability of data per country. Values in parentheses show the P values. PLHIV indicates people living with HIV; ART, antiretroviral therapy.
Correlation between average spending on care and treatment per PLHIV—adjusted (between 2009 and 2013) and AIDS death rate in 2013. Average spending on care and treatment for countries are calculated across a varying number of years depending on the availability of data per country. Values in parentheses show the P values. PLHIV indicates people living with HIV.
Trends in Expenditure by Program Area
In 2009, according to available data for 33 countries, care and treatment accounted for approximately 50% of total program spending. In 2012, the proportion of spending on care and treatment was 46% (based on data from 22 countries). The proportion of total HIV spending allocated for prevention programs and for program management and administration strengthening has increased marginally over a period of 4 years (2009-2012; Figure 5).
Spending trends by program area, 2009 to 2012. Of the 38 countries, data were available from 33 countries for 2009 (excluding South Africa), 29 countries for 2010, 33 countries for 2011, and 22 countries for 2012. Mgt indicates management; Admin, administration.
Trends in HIV Spending on Care and Treatment, ART Coverage, and AIDS-Related Deaths in Nigeria and Swaziland
Reported total HIV spending in Nigeria increased by 36% from US$420 million in 2009 to US$570 million in 2012. However, spending on care and treatment declined from US$200 million (48% of HIV spending) to US$190 million (33% of HIV spending) during the same period (Figure 6). Additionally, during 2009 to 2012, estimated AIDS-related deaths increased from 205 000 to 221 000 while ART coverage increased marginally from 9% to 15%. The year 2013 saw a decline in AIDS-related deaths to 210 000, while ART coverage increased to 20%.
Trends in HIV spending by program area, ART coverage, AIDS-related deaths, and new infections in Nigeria and Swaziland. ART indicates antiretroviral therapy.
In Swaziland, total HIV spending increased by 29% from US$75 million to US$97 million from 2009 to 2013. HIV spending on care and treatment increased from US$14 million (19% of HIV spending) to US$51 million (53% of HIV spending; Figure 6). The ART coverage among PLHIV has increased from 16% in 2009 to 49% in 2013. Between 2009 and 2013, the estimated AIDS-related deaths and new HIV infections have declined by 2500 (41%) and 4000 (29%), respectively.
Discussion
The new 90-90-90 targets, which build on the previous target to place 15 million PLHIV on ART by 2015, 16,17 were formally adopted by UNAIDS during the UN General Assembly Special Session in September 2014. The targets propose that by 2020, at least 90% of all PLHIV will know their status, 90% of them will be on sustained ART, and 90% of them will be virologically suppressed. Mathematical modeling suggests that achieving the 90-90-90 targets by 2020 will enable virtual elimination of the HIV pandemic by 2030 (defined as a ≥90% decrease in disease burden (including morbidity, mortality, and incidence) using 2010 levels as a baseline). 17 Achieving these targets will undoubtedly have a major public health impact. However, our study illustrates a serious lack of accurate financial data in the public domain. The data that were available highlighted a wide variation in the sources, level, and estimated impact of funding for care and treatment. These findings, the need to expand from roughly 41% to 81% ART coverage for PLHIV, and the current funding levels suggest that achieving the 90-90-90 targets will require a reexamination of the current resource allocation strategy.
The call for the new 90-90-90 targets came at a time when many countries, as well as PEPFAR and the Global Fund, are struggling to increase their funding levels. 13 –15 Each of the estimated 37 million PLHIV will eventually need ART immediately for their own health. 28 In this context, our study highlights the importance of increasing the focus and efficiency of available investments to ensure that PLHIV have access to care and treatment, in a timely fashion. This investment is critical not only for individuals living with HIV and their families, but it is increasingly recognized as integral to bending the HIV/AIDS-related morbidity and mortality curves toward zero.
Investment in HIV care and treatment will pay off as improved access to earlier HIV diagnosis and ART for PLHIV achieves reductions in AIDS-related illness and deaths, HIV transmissions, and health and other costs. 2 –9 Despite these clear benefits, only 15 of the LMICs we examined allocated more than 50% of their HIV spending to care and treatment, and many high-prevalence countries were well below this allocation level. We found a clear association between allocated AIDS response spending, ART coverage levels, and estimated AIDS-related deaths with countries that are investing more on care and treatment achieving better ART coverage and subsequently fewer estimated AIDS-related deaths. Although trend data are not available for most countries, the proportion of spending on care and treatment among the LMICs we examined does not appear to match the new science around the preventative and clinical benefits of earlier ART initiation. 3 –5 Because the 90-90-90 targets were only recently embraced by many UN member states, 16,17 they are likely only now being incorporated into national AIDS plans. However, the trend data suggest that current allocations in many countries will need to be significantly increased to reach the new targets and, therefore, the incorporation of 90-90-90 targets into national AIDS plans should be accompanied by increases in HIV spending on care and treatment.
There is also a significant variation in the allocation of HIV funds for care and treatment among countries. Arguably, this reflects previous WHO recommendations to “know your epidemic” and “tailor your response” accordingly. 29 However, a closer examination of the available data raises a significant concern, given that many of the high HIV burden countries in East and Southern Africa are spending less than 40% of their HIV funding on care and treatment. Of note, the proportion of spending on care and treatment is associated with per capita GDP but not whether the funding was from domestic or international sources. Although the impact on AIDS-related deaths and new HIV infections cannot be attributed to spending on care and treatment alone, the variability and disparities are starkly highlighted when comparing Nigeria with Swaziland. While chosen for different reasons and clearly quite different, their contrasting trends are influenced by their respective allocation of resources. Nigeria has a significant epidemic with 3.2 million PLHIV in 2013 and has actually decreased spending on care and treatment by 15%, despite a 36% increase in overall resources. Swaziland, reporting the highest prevalence rate in the world, has increased allocated resources 4-fold for HIV care and treatment. There is a stark contrast in important program and health outcomes such as ART coverage, estimated new infections, and AIDS-related deaths. The other worrisome aspect of the trend analyses is the lack of increase in expenditure for care and treatment over time, despite changing WHO guidelines and new HIV testing, ART coverage, and ART outcome (viral suppression) targets. Arguably the cost of ART has fallen dramatically over the past 5 years, suggesting that a static proportion of spending on care and treatment may translate into more PLHIV placed on ART for the same level of resources spent on diagnostic and drug commodities, for example. A closer country-specific look at the link between resources, prioritization of interventions, and impact would help to allay these concerns.
There are a number of important limitations to our work. First, despite considerable investments made in monitoring the global AIDS response, the data in the public domain regarding economic investments are incomplete and require cautious interpretation, particularly when examining trends. Little is known about the quality and veracity of the country reported spending data, and international and domestic funding are often used to provide different types of services within and outside government settings at different unit costs. The limited database does not allow us to discern the details of care and treatment spending which would help evaluate specific costs and impact. This is important when considering ways to more efficiently prioritize existing resources and the time lag between interventions and health outcomes. The bottom line, however, is that while there are strong concerns about the quality of the national expenditure data, these are currently the best data available. While countries do report their data, the lack of consistent reporting over time precludes optimal trends and multivariate analyses. The lack of data complicated our ability to match ART coverage data with spending data, and we thus had to use dated coverage information from 2013. HIV spending allocation codes are susceptible to variation, and funding is often used for a number of program interventions across categories. Reported HIV spending does not capture the contribution to the response from communities and individuals living with HIV and their families. The cost of the response also varies significantly, with some high-income countries in the North paying more than 1000 times the cost for annual ART when compared with lower income countries, some of that variability due to variations in GDP and/or diagnostic and drug commodity pricing at brand versus generic prices. Linking the reported allocation of resources with health outcomes is important but needs to be interpreted with caution, particularly as some of the outcomes such as AIDS-related deaths are derived from assumptions around ART coverage. Additionally, there are complex interactions between different prevention interventions including treatment, rendering causality difficult to determine. The data may not reflect recent prioritization shifts such as those seen in PEPFAR 3.0, which prioritizes high-impact interventions in and among high HIV burden geographies and populations to meet the 90-90-90 targets and achieve epidemic control. 30
Earlier diagnosis and ART initiation is essential to prevent immune deficiency, illness, death, and HIV transmission. 3 –5 In September 2015, the WHO issued new HIV treatment guidance that recommends ART initiation irrespective of CD4 count for all PLHIV. 27 International and national guidelines are increasingly focused on ensuring earlier HIV diagnosis and ART initiation; 14 countries now recommend offering ART after diagnosis irrespective of CD4 count. 19,27,31 Global implementation of these and other guidelines recommending earlier ART initiation could contribute to achieving the 90-90-90 targets and averting 28 million new HIV infections and 21 million AIDS-related deaths by 2030. 12 Increasing ART access will support reaching the 90-90-90 targets but will require improved focus, efficiency, and, in many settings, increased resources for care and treatment. However, with respect to the latter, current flat-lined trends in international financing suggest the global HIV community will need to do more with the same or even less resources through efficiencies and prioritization mechanisms. Without innovative means of service delivery including community testing and treatment delivery, decreasing the use of unnecessary laboratory tests such as CD4 counts for ART initiation, providing patients with 6 months of ART, and other efficiency gains, it will be impossible to reach and sustain the 90-90-90 targets. Although the science would suggest that with innovative delivery of ART and other prevention interventions, such as voluntary male medical circumcision and condoms, we have the means to control the HIV epidemic, it will be increasingly important to understand what resources are available and to make strategic investments in the right interventions for the right locations.
The 2015 WHO guidelines test-and-treat recommendation changes the global standard of care and should result in a major shift in strategy to expand access to ART to achieve the 90-90-90 targets and control the HIV epidemic. Our work benchmarks spending on HIV care and treatment in LMICs, highlights the relative lack of prioritization for treatment, and also starkly illustrates deficits in available data regarding resources for the global HIV response. The paucity of available and transparent financial data and the need to double global treatment access indicates that we will need to improve the ability to monitor and evaluate investments in the HIV response. There is a limited time window to achieve epidemic control before the number of PLHIV outstrips our ability to provide care, treatment, and other services. Just as ensuring access to treatment is the highest priority, guaranteeing an efficient use of the investment is urgently needed to maximize the individual, community, and public health impact of our collective response to the global HIV epidemic.
Footnotes
Acknowledgment
We would like to thank Brad Hersh for his careful review of the manuscript.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.