How can we evaluate the cost-effectiveness of health system strengthening? A typology and illustrations.

Health interventions often depend on a complex system of human and capital infrastructure that is shared with other interventions, in the form of service delivery platforms, such as healthcare facilities, hospitals, or community services. Most forms of health system strengthening seek to improve the efficiency or effectiveness of such delivery platforms. This paper presents a typology of ways in which health system strengthening can improve the economic efficiency of health services. Three types of health system strengthening are identified and modelled: (1) investment in the efficiency of an existing shared platform that generates positive benefits across a range of existing interventions; (2) relaxing a capacity constraint of an existing shared platform that inhibits the optimization of existing interventions; (3) providing an entirely new shared platform that supports a number of existing or new interventions. Theoretical models are illustrated with examples, and illustrate the importance of considering the portfolio of interventions using a platform, and not just piecemeal individual analysis of those interventions. They show how it is possible to extend principles of conventional cost-effectiveness analysis to identify an optimal balance between investing in health system strengthening and expenditure on specific interventions. The models developed in this paper provide a conceptual framework for evaluating the cost-effectiveness of investments in strengthening healthcare systems and, more broadly, shed light on the role that platforms play in promoting the cost-effectiveness of different interventions.

Vaginal microbicides save money: a model of cost-effectiveness in South Africa and the USA.

OBJECTIVE: To determine the hypothetical cost-effectiveness of vaginal microbicides preventing male to female HIV transmission. METHODS: A mathematical epidemiological and cost-effectiveness model using data from South Africa and the USA was used. The prospective 1-year-long intervention targeted a general population of women in a city of 1,000,000 inhabitants in two very different epidemiological settings, South Africa with a male HIV prevalence of 18.80% and the USA with a male HIV prevalence of 0.72%. The base case scenario assumes a microbicide effective at 55%, used in 30% of sexual episodes at a retail price for the public sector in South Africa of US$0.51 per use and in the USA of US$2.23 per use. RESULTS: In South Africa, over 1 year, the intervention would prevent 1908 infections, save US$6712 per infection averted as compared with antiretroviral treatment. In the USA, it would be more costly: over 1 year, the intervention would prevent 21 infections, amounting to a net cost per infection averted of US$405,077. However, in the setting of Washington DC, with a higher HIV prevalence, the same intervention would prevent 93 infections and save US$91,176 per infection averted. Sensitivity analyses were conducted and even a microbicide with a low effectiveness of 30% would still save healthcare costs in South Africa. CONCLUSIONS: A microbicide intervention is likely to be very cost-effective in a country undergoing a high-level generalised epidemic such as South Africa, but is unlikely to be cost-effective in a developed country presenting epidemiological features similar to the USA unless the male HIV prevalence exceeds 2.4%.