The Cost-effectiveness of Intervening in Low and High HIV Prevalence Areas in South Africa

The cost-effectiveness of intervening with a set of HIV/AIDS interventions in low HIV prevalence areas (LPA) and high HIV prevalence areas (HPA) in South Africa is analysed. The rationale for this analysis is to assess the suspected effect of interaction between the intervention and area of implementation, on cost-effectiveness. The paper used the Markov model, which tracked a cohort of patients over their lifetime in each area. Data on costs and health outcomes were collected from the literature, but the distribution of patients in health states at baseline and over time, were based on the patterns observed in the Actuarial Society of South Africa AIDS model (ASSA2008) projections, to depict these interaction dynamics. The effects of recent changes in guidelines of some interventions under consideration were assessed separately outside of modelling and sensitivity analysis conducted on all model parameters. In terms of efficiency, the study found it more cost-effective to intervene in LPA. However, to align efficiency with equity and ethical principles underlying HIV response, more than proportional resources should go into non-ARV based interventions in LPA, while more than proportional resources should go into non-ARV interventions in HPA.