Skip Navigation


Global Projects


Cost-effectiveness of public-private partnerships for HIV infection

Cape Town, South Africa


Scaling up access to antiretroviral therapy (ART) is a major challenge in Africa. Health care infrastructure in the public sector is variable, but everywhere in the region there is a shortage of skilled medical personnel. In many countries there are more medical practitioners in the private than in the public sector. A public-private partnership whereby public sector patients access private primary care doctors offers a potential solution to improve access to ART, but the cost-effectiveness of this approach needs to be analysed. Two public sector cost-effectiveness studies of ART have recently been completed in Cape Town. The largest private sector HIV/AIDS disease management program in Africa (approximately 40,000 patients from nine countries have been registered) is based in Cape Town. This program is providing support and data management for a PEPfAR-funded program that offers private sector primary care doctor care to public sector patients - this program has just commenced. We aim to undertake a cost-effectiveness analysis of ART in the private sector through an existing collaboration with the disease management program. Then the cost-effectiveness of the public-private program will be evaluated. These two models will be compared with the two public sector cost-effectiveness studies Private sector and public-private partnership cost-effectiveness will be analysed using the pre-ART period as the comparator. Costs will be measured from the provider’s perspective. Costs and effects will be discounted appropriately. Incremental cost-effectiveness ratios will be used as summary measures. Each cohort (public, private and public-private) will be matched for baseline CD4 count, HIV viral load, age and gender to ensure comparability. Intermediate outcomes (e.g. change in CD4 count and viral load by duration on ART) and final outcomes (deaths) will be measured in each cohort, and will be used to calculate the transition probabilities required to operationalize the Markov model. Uncertainty will be assessed through probabilistic sensitivity analysis. Monte Carlo simulated Markov models will be used to calculate lifetime costs, quality-adjusted life years and life years gained.


  • Start Date: 
  • End Date: