A33 Using serological and surveillance data to infer the introduction date and unobserved transmission dynamics of Zika virus in Fiji 2013–7

Zika virus (ZIKV) has been circulating in the South Pacific since 2007, and transmission in Fiji was first confirmed in 2015. To better understand the history and transmission dynamics of ZIKV in Fiji, we combined a transmission dynamic model with serological and surveillance data from Central Division, Fiji. A longitudinal population representative of seroepidemiological data were available from participants sampled in 2013, 2015, and 2017. In addition, ZIKV case reports were available from 2015 and 2016. Using a Bayesian approach, we fitted a transmission dynamic model with a seasonally varying transmission to these data. We also estimated the virus introduction date, given the effect this has on transmission dynamics as it interacts with the observed seasonal pattern of transmission. We found evidence that the virus was introduced in October 2013 (95% credible interval: April 2013–April 2014) and that the strong seasonal transmission pattern meant the virus persisted for several years with multiple waves of infection in consecutive years. It is important to corroborate this evidence against other work done in the same area. A phylogenetic analysis was performed on 5 ZIKV strains obtained from Fiji in 2015 and 2016, which were aligned with 33 E gene sequences from the Pacific, Americas, and Africa. This analysis showed evidence of virus persistence over multiple years in Central Division, Fiji. The estimated most recent common ancestor of the group isolated from Central Division was November 2013 (95% credible interval: March 2013–July 2015). Our modeling estimate is consistent with these results despite the very different methods being used. The availability of detailed case and serology data in an island outbreak setting, combined with mathematical models, presented a unique opportunity to gain crucial insights into these infections. Our analysis provides evidence that seasonal variation in transmission, combined with other co-circulating flaviviruses, means the timing of ZIKV introduction can have a major impact on outbreak transmission dynamics.