The goal of this project is to develop statistical and mathematical models of malaria transmission that will inform key features of program implementation, helping to optimize surveillance and control strategies at the community level. For this, we propose coupling, at the level of the malaria-affected district of Ifanadiana (south-east of Madagascar), accurate epidemiological surveillance of malaria cases, high resolution satellite environmental information, and longitudinal socio-economic and behavioural data, to gain a comprehensive understanding of local malaria dynamics that can inform program implementation in real time. We will assess whether the availability of this information helps improve local malaria control programs. By leveraging an existing healthcare delivery platform and robust data collection systems in Ifanadiana set by the healthcare NGO PIVOT, we will pilot innovative and reliable malaria decision-making tools that can be validated locally and scaled-up globally.
The specific objectives of this project are 1) to estimate the unobserved burden of malaria at the community level for improved surveillance. Our hypothesis is that more than one third of symptomatic cases are not captured by the district’s facility-based surveillance system due to financial and geographical barriers to access healthcare; 2) to integrate community-level predictions of malaria transmission into existing CHW workflows in Ifanadiana district for improved program implementation. We hypothesize that the use of routine predictions of malaria cases at the community level by CHW supervisors will improve case detection and will reduce stock-outs of RDTs and antimalarial drugs; 3) to inform implementation of additional control strategies that minimize malaria transmission through transmission model simulations. Our hypothesis is that implementation of malaria control activities based on a profound understanding of local transmission will lead to further reductions of malaria incidence than routine control strategies could achieve.