Why $10 billion for AIDS?

It is a memorable round number, a huge advantage for slogans like “where’s the $10 billion a year that is needed for HIV/AIDS?” But is it anything more than that?

This question was addressed at the International Conference on AIDS in Barcelona by Dr Stefano Bertozzi, an academic economist working in California and at the National Institute of Public Health in Cuernavaca, Mexico, in a session on “Financing the Global Epidemic”.

What emerged from the presentation is that there are at least two different but overlapping “$10 billion” figures under discussion, and there is a strong case for a larger sum to be spent on meeting global health needs, plus education and other action needed to reduce poverty.

The UN General Assembly’s Declaration of Commitment committed the UN’s members to expanding prevention efforts and increasing access to care and support. This presents a challenge, in estimating what resources are actually needed. Dr Bertozzi had worked with a group of colleagues in preparation for UNGASS to try and answer this question.

The estimates they made are based on scaling up responses by 2005 without significant strengthening of existing infrastructure, apart from providing some training, and sorting out drug distribution systems.

Estimated cost of scaled up care and prevention, by region/millions of US$

Africa - care: 3070 prevention: 1560

South SE Asia - care: 670 prevention: 1440

East Asia - care: 89 prevention: 810

Latin America - care: 550 prevention: 590

E Europe C Asia - care: 20 prevention: 250

N Africa M East - care: 50 prevention: 160

Total 9.2 billion - care: 4440 prevention: 4800

The care figures are based, firstly, on costs which tend to be concentrated in the last two years of life before someone dies, plus the costs of treatment through additional years of life gained when people are treated with antiretrovirals. This means that care costs vary between regions following the number of people with AIDS.

Care was broken down into palliative care, OI treatment (some of which, such as TB treatment, didn’t depend on knowing people’s HIV status), increasing the proportion of people knowing their HIV status, providing OI prophylaxis, ARVs.

The economists looked at what access people have now, then estimated the maximum rate of change that could be achieved in a particular country’s healthcare system. Some measures of this can be seen by looking at their past record, e.g. of achieving coverage with EPI vaccines (for children).

Care costs therefore vary with the ability of healthcare systems to provide antiretrovirals. In Asia, this capacity is greater than it is in Africa, increasing costs relative to the number of people with AIDS. The projection is based on estimates, country by country, of how rapidly the number of people on treatment could be scaled up, in effect, if the drugs were available free of charge. This is where the target of three million people on treatment by 2005 has come from, and the cost of drugs (at reasonably discounted prices) to meet this target has then been included in the figures set out above.

Prevention is entirely different in the pattern of spending, because it is more closely related to the total population of the countries concerned. Also, while scaling up care MUST be based on the healthcare system, prevention is not so limited. Some things – such as changes in the school curriculum – can be scaled up very rapidly, others are much harder to achieve.

What emerges is an estimate of what each country could productively spend every hyear on a basic package of prevention, care and support of orphans.

It is important to understand that the figures given above include both national and international resources. It also tries to include both public and private sources. Prevention has to be from public resources, much of care costs are in fact born by families.

These estimates are based on a great many assumptions about the relationship between costs from pilot studies and full-scale provision. Therefore, the group did study the impact of changing those assumptions within reasonable limits (a “sensitivity analysis”). The effect was to vary the estimates between 7-10 billion, but not 2 to 200!

What $10 billion won’t do

This figure is NOT an estimate of the resources needed to care for all people living with HIV and AIDS or to prevent all new HIV infections in 2005.

The figures also do not include any funding for expanding capacity of existing infrastructure (e.g building new hospitals, training new doctors and nurses). Yet it is already clear from countries such as Botswana, that this will be essential if non-AIDS-related health services are to be sustained.

An alternative and more realistic estimate for the Commission on Macroeconomics and Health arrived at a figure of US$13.6 to 15.4 billion should be spent on HIV/AIDS prevention and care annually in 2007, in selected low and middle income countries, INCLUDING money to grow infrastructure. A global figure on this basis would obviously be higher still.

The figure of $10 billion is also not a direct estimate of what should be in the Global Fund to fight AIDS, TB and Malaria. Some regions can largely fund their own response, having smaller epidemics, and more resources – so their need for international assistance from the Global Fund will be very different. Overall, between a third and a half can be met from national funds, but this falls to between 18% and a third for Sub-saharan Africa. So if the figure of $10 billion were the overall target the need for the Global Fund’s HIV/AIDS component is more like $3 to 4 billion.

It is hard to be sure how much could be saved with better prices on antiretrovirals. The economists took an idealised view in which the poor pay less than the rich. Assumed 350 dollars for poorer countries, and 4000 dollars for wealthier middle-income countries. Unfortunately, the emerging pattern seems to follow negotiating power rather than disease burden or level of poverty. Brazil with a GNP per capita twice that of Ecuador is paying a third of the price for antiretrovirals.

A global or regional developing country buying cooperative could make pricing more rational than this. Prices should be related to epidemic size as well as national income.

The model also didn’t include any allowance for inefficiency, corruption and so on.

So in the real world, these estimates are underestimates rather than overestimates.

A couple of days ago, the New York Times pointed out in an editorial that in USA, Europe and Japan agricultural subsidies are around $350 billion a year. These cost the developing world about $50 billion a year, according to the UN Development Programme – about the same as the sum total of development assistance provided by the wealthiest countries to the poorest.

The bottom line is that the resources needed to make a major change by 2005 are not there yet, but are more than reasonable, urgent, and highly feasible. An almost negligible change in agricultural subsidies would fund the entire amount.

In discussion, one of the questions raised was whether collective negotiation among countries on a regional basis (for example, within the Latin American region, the Caribbean or the member states of the African Union) would be compatible with other measures to limit costs such as local production and/or importation of low-cost generic drugs. Dr Bertozzi thought there was no reason why both strategies could not co-exist. Also, groups of countries could negotiate for drugs from different companies to be combined in fixed doses in a way that individual countries could not. Such collective arrangements would also act as a barrier to corruption by cutting the scope for private negotiations between individual governments and companies.