UC San Diego

Background: Aedes aegypti-borne diseases, including dengue, chikungunya, and Zika viruses, are increasingly important public health problems. Detecting and monitoring the transmission of arboviruses is critical for prevention and control activities. This dissertation research aimed to identify environmental, geospatial, and sociodemographic risk factors for arboviruses and their vectors in Guatemala, and compare vector and human surveillance strategies to detect local arbovirus transmission in Puerto Rico—sites of recent large arbovirus outbreaks.

Methods: This dissertation includes three studies examining arbovirus and vector risk factors and surveillance strategies. Study one was a retrospective case-control study using data from a prospective public health surveillance system in hospitals and clinics in Guatemala and logistic regression to examine whether arboviral infections were associated with indicators of household air pollution. In study two, we used data from human and mosquito surveillance systems in Puerto Rico to develop a simulation model to compare these surveillance systems for detecting and monitoring Zika virus activity. Study three included data from two cross-sectional household surveys in Guatemala and used generalized linear and generalized additive models to assess whether household environmental and geospatial factors were associated with immature mosquito abundance.

Results: For study one, arboviral infections were inversely associated with cooking with firewood in the main house, on an open hearth, and ≥5 times per week. In study two, both vector and human surveillance strategies effectively identified transmission in simulated high Zika virus transmission scenarios. In simulated low incidence scenarios, vector surveillance had higher sensitivity than human surveillance and that sensitivity increased with more traps and tests. In study three, proximity to paved roads and other houses/structures was predicted to be associated with greater immature mosquito abundance. Households with low and high household infrastructure had fewer larvae/pupae than households in the middle range.

Conclusion: Better understanding of factors defining geographical distribution of arboviral vectors may allow for improved targeting of vector surveillance, prevention, and control measures in areas considered at higher risk for arbovirus transmission. Virological surveillance in mosquitoes may improve sensitivity for arbovirus detection compared to human surveillance, but resource availability is an important factor when considering the most effective approach.