Modeling the control of West Nile virus using mosquito reduction methods, vaccination of equids, and human behavioral adaptation to the usage of personal protective equipment

West Nile virus (WNV) is a mosquito-borne virus of the genus Flaviviridae circulating between mosquitoes and birds, while humans, equids, and other mammals are dead-end hosts. Several preventive measures are recommended to reduce the WNV burden among different hosts. In this work, we develop a mathematical framework for evaluating the theoretical effectiveness of various WNV control methods in Germany. We consider mosquito reduction methods such as the physical removal and destruction of potential mosquito breeding sites, the use of larvicides, and the use of adulticides. We also evaluate the usage of personal protective equipment (PPE) that aims to reduce human-mosquito contact. Adopting PPEs may not happen instantly due to different perceptions, social influence, and the perceived inconvenience and/or frustration that come with using PPEs. Thus, we model a dynamic adoption of PPEs by considering the perceived risk of infection, perceived inconvenience of using PPEs, and the imitation dynamics due to social influence. Furthermore, the model captures vaccination of equids. We formulate and study an optimal control problem, where mosquito controls are temperature-dependent and the decision to start or stop applying the control methods is influenced by the changes in temperature. The optimal control model supports the development and seasonal timing of cost-effective mosquito control methods. For example, results from the optimal control study show that mosquito control efforts in Germany should be initiated during early spring and stopped at the end of June or early July, during the first year of control, to avoid overuse and unnecessary costs. Finally, we developed a Shiny app that allows users to test how different combinations of interventions could reduce WNV cases.