74 Zika in the United States: Dynamics, Transmission, and Strategies for Control

Abstract

Background: The current Zika epidemic is the largest recorded, and the frst to occur outside of the areas of Africa and Asia. Its transmission continues where Aedes mosquitos are established. Continued transmission anywhere requires nations to remain vigilant and informed to prevent and control local Zika virus transmission and its associated outcomes. Methods: We developed a stochastic in-silico model to simulate Zika virus transmission given a single index case into every county in the United States and Puerto Rico. Our model utilizes high-resolution Aedes risk maps, and accounts for temperature dependent viral and vector replication and life-stage dynamics. County-level birth seasonality was ft to monthly birth cohort data from every county to calculate expected number of frst-, second-, and third-trimester exposures across time and space. We also assess control strategies, such as adulticides that reduce Aedes survival, reductions in human-vector interactions that reduce mosquito biting rates, and vaccination. Results: Within the United States, outbreaks of Zika virus will be limited primarily to introductions during the summer months, and the Gulf Coast regions and Puerto Rico. While a majority (75%) of US counties demonstrated at least minimal potential for transmission, the vast majority had only sporadic and single transmission events. Among simulations with at least a single transmission event, 65% of counties (and 96% of counties outside of the Gulf Coast) had median outbreak sizes of <10. In the continental US, Harris County, Texas, had the largest epidemics, with median epidemic size across simulations of 12,200 cases (IQR 4,562-25,980) including 88 (IQR 36-197) frst-or second-trimester infections. In Miami-Dade County, Florida, where local transmission occurred throughout 2016, our results are in strong agreement with those reported by the Centers for Disease Control and Prevention (214 cases simulated 244 reported). When human-vector contact rates were reduced, transmission was stable until contact rates were driven below 50% of baseline. Probability of transmission from an index case was inverse-linearly proportional to mosquito adulticide use (% coverage) while outbreak size had a log-linear relationship. In regions with high transmission, vaccination was most effective at preventing initial transmission when coverage met or exceeded 75%. Incidence fell log-linearly with vaccine coverage. Conclusions: In the context of limited resources, a watchful-waiting or control strategy might be an optimal approach for local public health agencies, with resources deployed once a transmission event is detected, rather than attempt to prevent any initial transmission from an unknown index case. However, in high-risk areas, such as high-density population centers within the Gulf-Coast region, a prevention-oriented approach may be prudent as the fast pace of transmission following initial introduction could outpace even well-planned control strategies.