1st Health Economics Antiretroviral Roundtable

A group of southern African health economists and public health specialists involved in economic evaluations of antiretroviral therapy (ART) programmes met in Durban on 22nd September 2004. The meeting was organised by ECI KZN PLUS in collaboration with HEARD and the Harvard AIDS Institute. The aim of the meeting was to identify and discuss methodological issues related to economic evaluations of ART programmes and to identify priorities for health economics research, and particularly economic evaluations, in southern Africa.

Several health economists and public health specialists presented work they have done in the economic evaluation of HIV interventions, and particularly ART. Several methodological challenges in performing economic evaluations of HIV interventions in the southern African context were highlighted at this meeting. These include:

• Difficulties in valuing the opportunity cost of staff members in a pilot project setting and in an environment characterised by widespread staff shortages
• Costing of opportunistic infections is problematic as most patients are suffering from multiple indistinguishable opportunistic infections at once.
• The difficulties experienced in obtaining quality of life valuations and the appropriateness of valuing death as the worst possible health state, given religious and cultural beliefs about death in African culture.
• Basing cost estimates on treatment protocols can be misleading as medical practice sometimes departs widely from these protocols.


The future number of patients on ARV, loss to follow-up and changes in drug prices are the key uncertainties in modelling costs and effectiveness of interventions. Uncertainty in estimates of model parameters could be lessened through better sharing of findings among South African researchers. Poor data and the lack of good longitudinal data were identified as key challenges in performing economic analysis.

A set of research priorities for the economic evaluation of HIV interventions, especially ARVs was agreed on. These were: Refinement of cost-effectiveness input data; Research into equity issues surrounding the ART rollout in the region; Establishing the cost-effectiveness of different models of delivery of ART (issues of technical efficiency in ART delivery); Research into health care finance: resource allocation, budgeting and planning and Research into the impact of interventions at a provincial/national level, including the impact on the rest of the health care system.

Keynote Presentation: ?“Do we know what works??”

In his keynote address, Dr Robert Greener, Economics Advisor for UNAIDS showed that we know too little about the effectiveness of many standard HIV interventions and that we have only an approximate idea of what these interventions cost. Key decision-makers need better information than they currently have if they are to commit public resources to HIV interventions. We need better information on which treatment combinations are most effective; what proportion of people on ART fail; how long it takes for most patients to fail; what the average extension to life is on ARTs; what affects survival on ARTs and what parallel actions are necessary to enhance effectiveness. These questions have profound economic policy implications as they help us to estimate future treatment costs, the costs of more effective programmes, and to decide which programmes to support. Greener argued that although aggregate resources are not the key problem in fighting HIV/AIDS, it is vital to improve the empirical cost estimates, as this information is the key to making informed policy decisions regarding the allocation of scarce health care resources.

In order to generate robust cost projections, we need better information about the determinants of cost and how they might change; better estimates of current population size and current and future coverage rates. We also need data on issues that limit programme and absorptive capacity (the ability to spend effectively). Health service delivery, policy and managerial capacity, physical infrastructure, poor co-ordination, stigma and discrimination are some of the key factors limiting programme capacity in many countries. Economic evaluations should try to address these bottlenecks. Economic evaluation, is a vital tool for policy-makers and planners, who are consistently faced with difficult choices relating to the allocation of limited resources amongst competing priorities, and are in urgent need of better information.

Chris Desmond: Costing ART programmes in KZN:

Chris Desmond outlined the model developed by ECI KZN PLUS and Harvard AIDS Institute to cost a sample of HIV/AIDS programmes in KZN at current levels and to estimate the costs of expanding these programmes as patient numbers increase. Chris Desmond explained the methodological problems encountered in conducting the research.

Dr Marionette Holmes: Economic Evaluation of ART in Botswana.

Marionette Holmes outlined the economic evaluation of ART in Botswana, which she is currently leading. The study is a detailed cost-effectiveness analysis (CEA) performed from a government perspective, which is attempting to capture health state valuations in Botswana. Costs and outcomes are modelled over 10 years, using a Markov simulation and data is being combined from several sources in Botswana. Dr Holmes?’ presentation is not included on this website as the project is still underway.

Dr Andrew Boulle: Projecting the cost of ART in SA: approaches and uncertainties.

Andrew Boulle and Susan Cleary presented different facets of the CTAC and Khayelitsha cost-effectiveness studies. Andrew presented on the approaches and uncertainties involved in costing ART in SA, with particular reference to the Khayelitsha study. He also presented the Markov model used to estimate costs and utilisation in this study.

Susan Cleary: Six steps to the evaluation of ART

Susan Cleary used a six step approach to evaluation to outline the evaluation of three different ART programmes (the Khayelisha pilot site, CTAC cost-effectiveness analysis being conducted with Motasim Badri and Robin Wood and the CIPRA cost-effectiveness analysis being performed with Debbie Muirhead). Susan Cleary worked through these three cost-effectiveness studies showing the different approaches used to each step in each study. Andrew Boulle and Susan Cleary shared a few of the methodological problems experienced in the evaluation of the Khayelitsha ART pilot.

The presentations from this meeting can be downloaded on the righthand side of this page