Big city, small world: Density, contact rates, and transmission of dengue across Pakistan.

Moritz U. G. Kraemer; T. Alex Perkins; Derek A.T. Cummings; Rubeena Zakar; Simon I. Hay; David L. Smith; Robert C. Reiner

Model selection was performed using backwards selection. Two base models were 184 investigated. First, a climate only model was created using only the climatological and climate 185 suitability covariates. Second, a "full" model using the density-dependent covariates as well as 186 the climatological covariates were combined into a single model which was then subjected to 187 backwards selection. For both models the mixing coefficient was initially set equal for each 188 district and once a final model was arrived upon, the mixing coefficient for Lahore was allowed 189 to vary separately from the other coefficients. All model fitting was conducted using R (R Core To explore the potential significance of spatial variation in mixing parameters, we conducted an 197 analysis to probe the inherent mathematical tradeoff between the mixing parameter and the 198 density-independent transmission coefficient . Specifically, to answer the question, what 199 difference in local transmission would be necessary to account for a difference in mixing while 200 achieving identical transmission dynamics. To explore this, we used eqn. (1) to establish: 201