BACKGROUND: In July 2016, Florida Department of Health identified the first autochthonous Zika virus infections in the continental United States. Since then, over 250 cases have been identified. Zika virus infection results in asymptomatic infections or mild illness for the majority of individuals, increasing the potential for more human movement while viremic. Identification of potential exposure locations is important to ensure proper response and control measures are put in place.
METHODS: Interviews of Zika cases and persons under investigation (PUI) included the collection of addresses visited within two weeks of symptom onset or sample collection date if asymptomatic. Addresses were mapped using ArcMap 10.3.1 software and were stratified by status (case/PUI), type of location (home/work/other), and onset date. Land use data were obtained from the 2015 Department of Revenue dataset. Addresses designated as the most likely source of exposure were determined by case clusters, reported mosquito bites, and time spent outside. Buffers of 150 meters, the Aedes aegypti flight range, were placed around the addresses. Cases identified in the buffer could trigger additional active surveillance. To more rapidly identify focal areas of transmission, newly identified PUIs were also mapped.
RESULTS: Addresses for over 300 individuals and 850 locations were mapped. On average, 2.9 addresses were mapped per person (range 1-19). Half of the cases were linked to one of three active transmission areas in Miami; however, a potential exposure location could not be identified for 71 cases. Half of the addresses were “other,” followed by home (33%), and work (18%). However, designated exposure locations involved home (39%), work (38%), and less frequently “other” (23%). Based on land use data, most case/PUI addresses within the active transmission areas (270 addresses on 161 parcels) were associated with businesses (58%; primarily restaurants, hotels, and retail locations) and 30% were residential locations. Overall, 136 people shared a common location (56 addresses total) and 120 additional addresses were located within 150 meters of another location. Most of these (113) were located in active transmission areas.
CONCLUSIONS: Mapping combined with epidemiologic data facilitated identification of case clusters and linkages to areas of active transmission. However, it may be difficult to determine exposure locations with highly mobile individuals. “Other” addresses provided additional data but identified fewer likely exposure locations. Focusing on “other” locations with increased potential for mosquito exposure may be more efficient. Further analysis of case location data may identify additional variables that can be used in risk models.