Cost-effectiveness of a malaria control programs in sub-Saharan Africa: analysis of uncertainties using a stochastic individual-based simulation model

Abstract

Background The first vaccine against P. falciparum malaria is close to licensure, having shown moderate levels of protection in clinical trials. Several analyses have shown the vaccine to be likely to be an eff ective and cost-eff ective addition to currently used control strategies. We previously used individual-based computer simulations (implemented in the openMalaria simulator) to analyze uncertainties in the predicted cost-eff ectiveness of introducing a malaria vaccine into the expanded programme on immunization in sub-Saharan Africa. We now extend these studies to include insecticide treated nets, a currently existing intervention. In addition, we also address model uncertainty in the current analysis. Materials and methods We used techniques of probabilistic sensitivity analysis, involving randomly sampling the parameter vectors, to analyze the contributions of the diff erent sources of uncertainty to the predicted cost-eff ectiveness. One specific aspect of these analyses of uncertainty is quantification of the value of acquiring additional information on these parameters by computation of expected value of perfect information (EVPI). Results Among the most important predictors of the cost-eff ectiveness of a control program are the cost of the intervention program and the transmission intensity at the time of the start of the program. EVPI is shown to be substantial, and in particular the accrual of up-to-date information on local endemicity would seem an efficient way to inform decisions about local deployment. Conclusions Probabilistic sensitivity analysis and value of information analysis using computer simulation models provide a powerful way to identify data gaps hindering rational resource allocation in malaria control.