Spatio-Temporal Patterns of Dengue Fever in Cali, Colombia

Dengue fever is an arboviral disease typical of the tropics that can be life-threatening and if not controlled properly may result in an epidemic. The absence of an effective vaccine makes strategies to prevent the virus transmission the most effective means of control. The planning of such strategies, however, is difficult due to the constant movement of individuals and mosquito host (Aedes aegypti). In this paper, the spatial and temporal relations that might exist between infected individuals during a dengue-epidemic year are explored. This research is motivated in that a deep understanding of potential transmission patterns between individuals might lead to a better design and planning of control strategies. A GIS-based Health Exploratory AnaLysis Tool (HELP) is used to compute space-time relationships by means of spatial K-function, kernel density, space-time K-function and linking pairs of cases within significant time and space intervals. Significant clustering was observed at a scale of 50 meters and 750 meters, respectively while temporal significance was determined at two days and five to eight days. While an increase of cases occurs in the months following severe droughts due to an El Niño phenomenon, the location of clusters remains relatively stable. These are observed near areas where potential habitats for the mosquito exist such as storm drains, hard surfaces where water accumulates (e.g., vases, containers), but also in poorer neighborhoods. The results from the spatial analysis provide valuable information for health care managers to take preventive actions at the municipality level.