Study sites and design.

This collaborative study was conducted in two geographically distinct regions; namely, Nicaragua (Central America) (Figure A1.1) and Ecuador (South America) (Figure A1.2). In Nicaragua, we worked in two districts within the capital city of Managua: District 2 (classified as urban) and District 3 (classified as urban-rural). The study focused on the catchment area of one public sector health center (District 2) and two health posts (District 3). Together, these districts comprise over 340,000 inhabitants and an overall population density of 3,800 people/km² (^{INIDE, 2021}). In Ecuador, our research focused on six communities (Borbón (urban-rural), Colon Eloy, Santa Maria, Santo Domingo, Maldonado and Timbiré (rural) located in the northwestern province of Esmeraldas. Based on our study census, the population in the study communities is ~20,000 individuals. These communities have a combined population density of 41 people/km² (^{Zambrano, 2023}). Both districts in Nicaragua and the communities in Ecuador encompassed low-income neighborhoods, with varying degrees of urban and rural characteristics, as well as differential access to healthcare services. The overall climate in both regions is tropical with marked rainy seasons (Nicaragua: June-December; Ecuador: March-July) (^{Thorsen, 2015}) that overlaps with the dengue transmission cycle.

The entomological surveillance programs established for both sites are part of ongoing community-based cohort studies evaluating arboviruses in each region. In Nicaragua, we randomly selected a subset of 500 households from each district (of 2,166 and 1,000 households in the parent cohort study of District 2 and 3, respectively) and ~50 key sites (i.e., tire shops, cemeteries, and schools) from each district. This selection was done using a framework for patches of risk based on landscape features. In Ecuador, an average of 30% of all community households were randomly selected. The number of households varied from 356 to 2,880 depending on the community, for a total of 546 households surveyed.