A Demonstration of the Use of SaTScan to Identify Spatio-Temporal Clusters of Salmonellosis

BACKGROUND:  Public health surveillance is a vital ongoing process for understanding the changing patterns of chronic and infectious diseases. Surveillance data can be used to identify disease outbreaks using statistical analysis tools such as SaTScan. There are several useful features in SaTScan that can be used for identifying the space-time patterns of disease. SaTScan can also be used in an automated disease surveillance system.

In this live demonstration, basic features of SaTScan to identify space-time clustering will be demonstrated using Salmonella data. The software, training handbooks, and sample data will be made available on the SaTScan website to facilitate learning how to use the tool.

METHODS:  Salmonella data were extracted from the New York State electronic clinical laboratory reporting system to obtain the diagnosis or onset date and ZIP Code of residence of each case. Each case was assigned a latitude and longitude based on the population-weighted centroid of the ZIP Code. The data were masked to ensure patient confidentiality for demonstration purposes. The data are imported into the SaTScan software, using a variety of data input formats.

RESULTS:  This demonstration shows how to work with SaTScan’s graphical user interface to analyze reportable disease data using the prospective space-time permutation scan statistic. SaTScan output formats will be displayed as tables, graphs, and maps using widely available tools such as Excel, ArcMap, and Google Earth.

CONCLUSIONS: The improved features of the SaTScan software have made disease outbreak detection more efficient. Results can be easily displayed as tables, graphs, and maps. Training materials are available for interested public health practitioners.