West Nile Virus transmission risk from environmental, animal and human surveillance data

West Nile virus (WNV) is an emerging mosquito-borne zoonosis and a growing public health concern in the changing climate. WNV is maintained in an avian-mosquito transmission cycle and occasionally spills over into humans who are dead-end hosts. About 80% of WNV infections in humans are asymptomatic but around 1% develop fatal neuro-invasive disease. Because human-to-human transmission can occur via organ or blood transplant, WNV circulation is closely monitored in Europe during each season for blood safety. In the absence of a vaccine and antiviral treatment against WNV in humans, surveillance is the main form of disease control. As climate changes, we urgently need to better understand WNV hosts' responses to the environment to forecast, and so control, this disease. This interdisciplinary project combines computational and lab-work and aims at (i) forecasting WNV transmission from high-resolution environmental, entomological, ornithological, and human surveillance data collected in Italy; (ii) characterizing temperature dependencies in European Culex mosquito dynamics and (iii) updating a WNV human transmission risk map for Europe. In (i) we will estimate WNV prevalence in Culex mosquitoes from trap data and investigate its association with ornithological and environmental data collected weekly over 2013 - 2021 in Emilia-Romagna (Italy). We will use a suite of statistical approaches including spatiotemporal Bayesian mixed-effects models, machine and deep learning applied to observed climate data and, in a separate analysis, past weather forecast data. These data are provided through an existing collaboration between Imperial College London, IZSLER (https://www.izsler.it/), ARPAE (https://www.arpae.it/en) and FEM (https://www.fmach.it/eng). In (ii) we will fill current knowledge gaps on the effect of temperature on European Culex mosquito traits by conducting lab experiments to characterise blood-feeding behaviour (e.g., the duration of the gonotrophic cycle, the period between a blood meal and oviposition and the biting rate) from controlled temperature conditions. In (iii), we will use the relationships obtained from (ii) and other covariates of interest from a literature review to develop mathematical and statistical models linking observed historical time series of human WNV cases reported in Europe by ECDC (https://www.ecdc.europa.eu/en) with spatially and temporally matched climate and environmental data (https://pearselab.github.io/areadata/), to generate an updated map of WNV human transmission risk across Europe.