Lower than expected effective human-mosquito contact may regulate dengue virus transmission in high-suitability low-incidence environments: A plausible explanation for limited dengue outbreaks in Florida, U.S.

Dengue, a vector-borne disease, puts at risk of infection nearly four billion people worldwide. Spread primarily by Aedes mosquitos, suitable mosquito habitat favoring survival, population maintenance, and an optimal extrinsic incubation period is required for successful transmission. Many areas, however, fall within the geographic range of the mosquito vector and favorable climatological conditions but experience only sporadic and short-lived outbreaks. Here, we characterized local dengue virus transmission in Florida, United States, where transmission is limited despite frequent travel-associated cases and high climate-based estimates of the basic reproductive number. To reconcile the difference between expected R0 values and the observed transmission, we tested four transmission scenarios based on ecological assumptions on mosquito vector biology, human-mosquito contact, and temperature profiles. Specifically, we sought to understand the extent to which differences in effective contact between humans and mosquitoes may modulate the degree of expected transmission and assess variation between counties. Furthermore, we investigated possible drivers of differential contact rates between counties in the context of socioeconomic, demographic, and environmental variables. We found that estimated human-mosquito contact rates across Florida were lower than those reported from endemic settings or estimated based on laboratory experiments, resulting in smaller outbreaks than expected given vector biology and local climatological conditions across the state. Associated explanatory variables moderating the degree of contact included population density and the built environment. The expected number of secondary cases generated using updated effective contact rates was in line with the existing epidemiological and genetic evidence, suggesting that low contact rates limit transmission from imported cases such that most result in no further transmission. These findings also imply that extended transmission risk is only present in southern Florida due to frequent travel-associated cases. The analytical framework presented here offers a new modifiable way to estimate the contact processes underlying dengue virus transmission in high-suitability low-incidence environments, which can be applied in other settings at the margins of endemicity, providing interpretable epidemiological insights instead of relying on traditionally used theoretical R0 values inferred from unmeasured entomological estimates.