Using Space-Time Pattern to Detect Zika Virus Clusters in Miami-Dade County—2016

Wednesday, June 7, 2017: 10:30 AM
400B, Boise Centre
Guoyan Zhang , Florida Department of Health in Miami-Dade County, Miami, FL
Isabel Griffin , Florida Department of Health in Miami-Dade County, Miami, FL
Anthony Llau , Florida Department of Health in Miami-Dade County, Miami, FL
Jennifer Lawrence , Florida Department of Health in Miami-Dade County, Miami, FL

BACKGROUND: Following the first travel-related Zika virus case reported by the Florida Department of Health in Miami-Dade County (DOH-Miami-Dade), a web-based interactive GIS map was developed to monitor Zika virus spread within the community. In July 2016, DOH-Miami-Dade reported the first locally-acquired Zika virus outbreak in the continental United States. In response to the appearance of local transmission, space-time patterns in desktop GIS were applied to evaluate the geographic location and scope of local Zika virus transmission in Miami-Dade County.

METHODS:

Zika virus data is retrieved daily from Merlin, the Florida Department of Health Epidemiology Disease Surveillance System, and transferred to the DOH-Miami-Dade GIS server. The interactive GIS map presents the distribution of travel-associated and locally-acquired cases throughout the county, which allows investigators to examine the possible linkage between travel-associated case density and the appearance of local transmission. The space-time pattern has been analyzed on weekly intervals by onset date. The distance interval was set at 150 meters (the flight range of the Aedes aegypti mosquito) and 1-square mile (the Centers for Disease Control and Prevention (CDC) recommended Zika-transmission boundary) in order to compare the best distance for detecting Zika clusters.

RESULTS:

Seven locally-acquired cases were first identified in week 27 (beginning of July). The count of locally-acquired cases reached its peak in week 33 with 31 cases. In comparison to travel-associated cases which peaked between the weeks of 31 and 33 with an average of 23 cases. The interactive web based GIS map shows travel-related and locally-acquired cases within Miami-Dade County; however, clusters of locally-acquired cases were located within the neighborhoods of Wynwood, Miami Beach, and Little River. Several hotspots of locally-acquired cases were detected by mining for spatial-temporal patterns, those hotspots did not overlap with travel-associated hotspots. Hotspots patterns did not change when the distance interval was adjusted between 150 meters and 1-square mile. The density of travel-associated cases did not predict areas of local transmission.

CONCLUSIONS: The use of an interactive GIS map to identify geospatial-temporal patterns to detect hotspots within the county was extremely beneficial to case investigations during this outbreak. These maps allowed investigators to timely identify high-risk areas. The maps also allowed public health officials to implement targeted approaches for public health Zika response and interventions. These maps also demonstrated that other factors may play a role in Zika virus transmission.