Modelling geographic variation in the cost-effectiveness of control policies for infectious vector diseases: the example of Chagas disease.

Few cost-effectiveness analysis (CEA) models have accounted for geographic variation in input parameters. This paper describes a deterministic discrete-time multi-state model to estimate the cost-effectiveness of vector control policies for Chagas disease, where implementation varies according to village characteristics. The model outputs include the total number of new infections, disability adjusted life years (DALYs) incurred, costs of associated healthcare, and total costs of the Ministry of Health's control policy for house surveillance and spraying. Incremental net benefits were estimated to determine Colombian villages in which it is cost-effective to implement the control policy. The robustness of these conclusions was evaluated by deterministic sensitivity analyses. The model should help provide a decision-support system to compare control policies and to allocate resources geographically.