Abstract
BACKGROUND: Detecting and monitoring the transmission of arboviruses such as Zika virus (ZIKV), dengue virus, and chikungunya virus is critical for prevention and control activities. Previous work has compared the ability of different human-focused surveillance strategies to detect ZIKV transmission in U.S. counties where no known transmission had occurred, but whether virological surveillance in mosquitoes could represent an effective surveillance system is unclear. OBJECTIVES: We leveraged a unique set of data from human and virological surveillance in Ae. aegypti during the 2016 ZIKV epidemic in Caguas, Puerto Rico, to compare alternative strategies for detecting and monitoring ZIKV activity. METHODS: We developed a simulation model for mosquito and human surveillance strategies and simulated different transmission scenarios with varying infection rates and mosquito trap densities. We then calculated the expected weekly number of detected infections, the probability of detecting transmission, and the number of tests needed and compared the simulations with observed data from Caguas. RESULTS: In simulated high transmission scenarios (1 infection per 1,000 people per week), the models demonstrated that both approaches had estimated probabilities of detection of greater than 78%. In simulated low incidence scenarios, vector surveillance had higher sensitivity than human surveillance and sensitivity increased with more traps, more trapping effort, and testing. In contrast, the actual data from Caguas indicated that human virological surveillance was more sensitive than vector virological surveillance during periods of both high and low transmission. CONCLUSION: In scenarios where human surveillance is not possible or when transmission intensity is very low, virological surveillance in Ae. aegypti may be able to detect and monitor ZIKV epidemic activity. However, surveillance for humans seeking care for Zika-like symptoms likely provides an equivalent or more sensitive indicator of transmission intensity in most circumstances.