BEST PRACTICES - SPACE-TIME DISEASE SURVEILANCE TOOLS

ClusterSeer (version 2.2)

ClusterSeer

Features

Local spatial statistics
Global spatial statistics
Scan statistics

Cumulative sum surveillance (Tango statistic)
Multiple comparisons procedures
Map, graph, table views (not linked)

Extensive help and tutorials
Sample datasets included

Comments

ClusterSeer has a broad suite of methods available for temporal, spatial, and space-time analyses. Although only one method (CUSUM) is labelled as a surveillance method, other space-time methods could be used for surveillance given the capacity for multiple comparisons corrections in ClusterSeer. The map display has very limited functionality and displays are not linked.

Selected case studies

Ward et al. 2008 - This study examines the utility of spatial analysis and geostatistical methods for analyzing spread of highly pathogenic avian influenza (HPAI) through a national poultry population. Outbreak data from Romania is used as a test case. A number of software packages are used to analyze data in this paper. ClusterSeer is used to for directional space-time interaction of outbreaks to estimate spread. This highlights one of the main advantages of ClusterSeer over software packages is that there a large number of methods implemented for spatial, temporal, and space-time analysis.

Grubesic et al. 2008 - Space-time interaction tests in ClusterSeer are described, and a case study into crime in Cincinnati demonstrates their application to crime patterns. The Knox test and Jacquez k-Nearest Neighbor Test are compared. While the focus of this study is crime, these same tests are often used for retrospective analysis of disease surveillance data. Confirmation of space-time interaction can indicate the presence of an infectious etiology.

Note: we were unable to find examples of ClusterSeer as part of a surveillance system.

Surveillance Objective(s)

Cluster detection, Data exploration

User Expertise

ClusterSeer offers a user friendly interface to a large suite of spatial methods. The majority are geared towards retrospective analysis of spatial and space-time clustering. Multiple views of the data allow for visualization and exploratory analysis of surveillance data. An extensive help system and tutorials are provided. Data can be input as text files or shapefiles .

Key Considerations

No batch mode for automated analyses
Not free software
Provides capabilities for all steps of analysis (not just statistical analysis)

More Information

Website: http://www.terraseer.com/products_clusterseer.php

Key Resource: ClusterSeeer Online Help http://www.terraseer.com/help/clusterseer/index.htm.

References

Grubesic, T., & Mack, E. 2008. Spatio-temporal interaction of urban crime. Journal of Quantitative Criminology 24, no. 3: 285-306. doi:10.1007/s10940-008-9047-5.

Ward, M.P., Maftei, D. Apostu, C. & Suru, A. 2008. Geostatistical visualisation and spatial statistics for evaluation of the dispersion of epidemic highly pathogenic avian influenza subtype H5N1. Veterinary Research 39, no. 3: 12 pages.