Abstract
Global emergence of arboviruses is a growing public health concern, since most of these diseases have no vaccine or prevention treatment available. In this scenario, vector control through the use of chemical insecticides is one of the most important prevention tools. Nevertheless, their effectiveness has been increasingly compromised by the development of strong resistance observed in field populations, even in spite of fitness costs usually associated to resistance. Using a stage-structured deterministic model parametrised for the Aedes aegypti--the main vector for dengue--we investigated the persistence of resistance by studying the time for a population which displays resistance to insecticide to revert to a susceptible population. By means of a comprehensive series of in-silico experiments, we studied this reversal time as a function of fitness costs and the initial presence of the resistance allele in the population. The resulting map provides both a guiding and a surveillance tool for public health officers to address the resistance situation of field populations. Application to field data from Brazil indicates that reversal can take, in some cases, decades even if fitness costs are not small. As by-products of this investigation, we were able to fit very simple formulas to the reversal times as a function of either cost or initial presence of the resistance allele. In addition, the in-silico experiments also showed that density dependent regulation plays an important role in the dynamics, slowing down the reversal process.