While most of sub-Saharan Africa is heavily affected by HIV, the epicentre of the disease is in southern Africa where the world's worst affected countries are located. Within this region, South Africa has the highest number of HIV-infected people, representing 17% of the global HIV-burden (UNAIDS, 2007). Approximately 6 million are HIV-infected and 0.5 million die annually in the absence of antiretroviral treatment (ART). While plans have been made to achieve equal access to ART by 2011, unless there are substantial increases in health-care spending, HIV-treatment will divert resources from existing interventions almost certainly at the expense of both efficiency and equity.
The use of mathematical programming techniques in the economic evaluation of health-care programmes was first suggested by in the context of cost-benefit analysis and later by with application to cost-effectiveness/utility analyses. It is only through the application of these techniques to the economic evaluation of alternative HIV-treatment strategies in South Africa that the full extent of the importance of affordability has come to light. This article thus argues both for the importance of mathematical programming as a technique and affordability as a criterion in priority setting.
As currently applied, resource allocation in HIV-treatment is ethically fraught. We reflect on two common approaches. The first involves the use of cost-effectiveness analysis combined with a threshold incremental cost-effectiveness ratio (ICER). The second is that followed by the South African government and many others in Africa concerned with increasing access to HIV-treatment: designing interventions and setting targets that are costed only subsequently. Both approaches lack explicit consideration of affordability.
Cost-effectiveness analysis is the economic technique most frequently employed in priority setting in developing countries, with a growing number of such evaluations of HIV-treatment recently . However, given that the implementation of a new ART policy will be likely to impose additional costs on the health-care budget in comparison with the status quo (i.e. the ICER is positive), the further interpretation of the results of such studies in terms of cost-effectiveness or technical efficiency is impossible. This is because the provision of the new intervention to the same number of patients will require an increased budget -it is therefore a question of allocative efficiency . Authors therefore attempt to build a bridge between technical and allocative efficiency through a discussion of the ICER relative to a willingness to pay (WTP) threshold per, say, QALY gained. However, the calculation of the threshold is complex; its use subject to criticism . Moreover, because many developing countries have yet to develop a threshold ICER policy, the use of cost-effectiveness results is at best complex, at worst of very doubtful value.