Evaluating Interventions for Plasmodium vivax Forest Malaria Using a Three-Scale Mathematical Model

Abstract. The rising proportion of Plasmodium vivax cases concentrated in forest-fringe areas across the Greater Mekong Subregion (GMS) highlights the importance of pharmaceutical and mosquito control techniques specifically targeted toward forest-going populations. To mathematically assess best-possible antimalarial interventions in the context of hypnozoite reactivation and seasonal forest migration, we extend a previously developed three-scale integro-differential equation model of P. vivax transmission. In particular, we fit the model to data gathered over a four-year period in Vietnam to gain insight into local P. vivax dynamics and validate the model’s ability to capture epidemiological trends. The calibrated model is then used to generate optimal schedules for mass drug administration (MDA) in forest-goers and gauge the efficacy of vector control techniques (such as long-lasting insecticide nets and indoor residual spraying) in forest-adjacent areas. Our results emphasize the dependence of optimal MDA timing on the demographics of the human population, the importance of interventions targeting the mosquito bite rate, and the need for efficacy in hypnozoite-targeting antimalarial drugs. Future work should reduce current constraints on the model (particularly the assumptions on homogeneous travel times in forest-going groups, full adherence to the antimalarial regimen, and the use of only one MDA round) while also introducing more robust algorithms for model fitting (in place of the current least-squares approach). Our framework for MDA optimization may also be extended to spatially heterogeneous settings outside of the GMS, including travel networks along the Amazonian border in western Peru.