Feature Head to head

Are we spending too much on HIV?

BMJ 2007; 334 doi: http://dx.doi.org/10.1136/bmj.39113.402361.94 (Published 15 February 2007) Cite this as: BMJ 2007;334:344
  1. Roger England (roger.england{at}healthsystemsworkshop.org)
  1. Health Systems Workshop, Grenada, West Indies

    Billions of pounds are being spent on the fight against AIDS in developing countries. Roger England believes that much of the money could be better used elsewhere, whereas Paul de Lay and colleagues argue that current spending is not enough

    HIV is receiving relatively too much money, with much of it used inefficiently and sometimes counterproductively. Data from the Organisation for Economic Cooperation and Development show that 21% of health aid was allocated to HIV in 2004, up from 8% in 2000.1 It could now exceed a quarter. Yet HIV constitutes only 5% of the burden of disease in low and middle income countries as measured by disability adjusted life years lost (DALYs),2 less than that for respiratory infections, perinatal conditions, or ischaemic heart disease. It causes 2.8 million deaths a year worldwide—fewer than the number of stillbirths, and much less than half the number of infant deaths.2 More deaths are attributable to diabetes than to HIV.3

    Even within sub-Saharan Africa, HIV funding is out of balance. HIV is the biggest single killer, contributing 17.6% of the burden of disease in 2001.4 But it received 40% of all health aid in 2004.56 Although incidence and prevalence have peaked in Africa,7 HIV aid to Africa increased by an average of $240m (£123m; €185) a year from 2001 to 2004.5 Global HIV expenditure increased by an average of $1.7bn a year in this period.8 The 2006 UN General Assembly high level meeting on AIDS called for annual HIV expenditure in low and middle income countries to rise from $8.3bn in 2005 to around $23bn by 2010.9 If, as now, aid constitutes a third of this expenditure, and if non-HIV health aid continues to increase at current rates, HIV would then claim half of all health aid.

    Are HIV interventions so cost effective that they justify this disproportionate spending? No, they are not. Costs per DALY averted are lower for immunisations, malaria, traffic injuries, childhood illnesses, and tuberculosis.1011 Much HIV money could be spent with more certain benefits on, for example, bed nets, immunisation against pneumonia, or family planning.

    An exceptional disease?

    Why has this happened? One factor surely has been the success of HIV lobbies and activists in promoting HIV as exceptional.12 In rich countries, HIV has become the crusade of the famous, fashionable, and influential. In high prevalence countries, HIV affects the middle classes more than the poor13 and is of more concern to them: middle class children do not die from pneumonia or malaria and middle class women do not die in childbirth.

    The exceptional status accorded HIV, and its excessive relative funding, has produced the biggest vertical programme in history, with its own staff, systems, and structure. This is having deleterious effects apart from underfunding of other diseases. These include separating HIV from sexual and reproductive health and creating parallel structures that constrain the development of health services. National AIDS commissions, country coordinating mechanisms, UN agencies, etc are tripping over each other for funds and influence.

    HIV is also affecting adversely the organisation of health services. Funding for prevention of mother to child transmission, for example, is producing separate structures rather than strengthening everyday antenatal care and maternal child health by making testing and prevention part of the routine work of nurses and midwives. Also, well funded HIV programmes attract staff from other health services, aggravating chronic shortages.

    Because HIV interventions are not integrated into health services, this excessive spending is not effective. Nevirapine or other prophylaxis is given for only 9% of pregnancies in women with HIV, and only 1.5 million people are receiving antiretroviral drugs.8

    What is all this money being spent on? Much of it goes on “multisectoral” activities and “mainstreaming” HIV into just about every social activity. These have become the emperor's new clothes of public health. The World Bank's evaluation notes: “projects are complex with many participants engaged in activities for which they have little capacity, technical expertise, or comparative advantage.”14 Much money is wasted in areas that reflect the interests of those on the AIDS industry payroll more than evidence. It could be more effective if used to strengthen public health, which already provides preventive interventions in other sectors, cooperating with local authorities and ministries. Moreover, claiming HIV as exceptional may have increased stigmatisation.15

    Health systems not diseases

    More health aid should be used to strengthen health systems that can integrate funding at country level and allocate it to evidence based priorities through effective delivery organisations, whether state or private. Sector wide approaches try to do this by pooling aid and government funding and spending it to an agreed plan.16 They should be more independent of government and more representative—able to drive a big shift to market mechanisms that create real incentives to deliver and use the mass media to empower poor consumers to influence demand and improve self medication.

    A global basket fund is needed to transfer sustainable and predictable funding to countries, avoiding the hugely unpredictable aid flows from fickle donors that make planning impossible.17 The Global Fund to Fight AIDS, Tuberculosis, and Malaria could abandon disease dedicated support to become this fund. Its participation in sector wide approaches would give a big boost to rational resource allocation. Improving health systems should form the platform for action and research now, transcending HIV and other disease-specific programmes.18

    Footnotes

    • ARTICLE
    • Competing interests: None declared.

    References