Arbovirus transmission dynamics in Fiji and the wider Pacific region

This research will perform statistical analysis and mathematical modelling of infectious diseases transmitted by mosquitoes in Fiji and other Pacific island countries. The diseases in question are transmitted by the Aedes genus of mosquitoes and are becoming an increasing global health problem. These viruses can all spread easily in the Pacific islands where these mosquitoes are abundant. These islands are valuable case studies because the small populations are relatively isolated from other countries so typically have lower levels of immunity to an invasive virus. As a result, they usually experience short epidemics compared to large, heterogeneous populations which can sustain endemic transmission and is more complicated to analyse. This project will use mathematical modelling to study the dynamics of arbovirus disease transmission in the Pacific, how these viruses interact and how disease transmission has changed over the past decade.

Several studies have examined the spread of individual diseases in the region so far but their application have had limited scope in application. This research therefore offers the opportunity to test current models and improve on their accuracy by combining data sources from across the region.

This research is focused on quantitative skills with the application of existing statistical modelling methods to new data, and the potential improvement of methods.

Historically, dengue virus (DENV) circulates in the region with each of the four serotypes taking approximately 12 years to reappear due to the gradual accumulation of susceptible individuals as the population turns over. But these cycles seem to be getting shorter recently leading to a greater disease burden for health systems across the region. In addition, there have been outbreaks of other arboviruses recently with Chikungunya virus (CHIKV) detected in 2011 and there was an outbreak of Zika virus (ZIKV) in Fiji, first detected in 2015. Applying mathematical modelling to data from these outbreaks can help identify riskier periods for transmission during the year and highlight geographic areas more vulnerable to the spread of arboviruses.

This project also affords an opportunity to develop and test methods that combine multiple data sources with mathematical models. Data on arbovirus outbreaks in the Pacific is complicated by overburdened public health surveillance systems, and the abundance of silent asymptomatic transmission that cannot be detected. To better capture the true burden of arbovirus transmission, longitudinal serological data were collected from participants in Fiji in 2013, 2015 and 2017. Pre and post epidemic sera can then be combined with surveillance data and viral sequence data to analyse the extent of infection on the island. This demonstrates the potential of mathematical modelling to combine multiple, disparate data sources to better understand how arboviruses transmit.

Keywords: arbovirus, mathematical modelling, antibody dynamics, serology, surveillance