Summary
Analysis
of disease data that has an implicit spatio-temporal component (such
as disease outbreaks, data generated by surveillance systems and specific
hypothesis-based veterinary field research) is a foundation of veterinary
epidemiology and preventive medicine. Components of this process include
exploratory spatial data analysis (finding interesting patterns), visualisation
(showing interesting
patterns) and spatial modelling (explaining interesting patterns). Spatio-temporal statistics and
tests are valuable when adding precision to qualitative verbal descriptions,
facilitating the comparison of distributions and drawing attention
to characteristics unlikely to be noticed by visual inspection. Quantifying
spatio-temporal patterns is important for understanding how disease
phenomena behave. The application of a range of spatio-temporal statistics
is illustrated by exploratory spatial data analysis and visualisation
of the 2002 outbreak of West Nile virus encephalomyelitis in Texas
equines. This large outbreak (1 698 reported cases) consisted
of both point (latitude, longitude) and polygon (Texas counties) spatial
data
with a time component (reported date of onset of clinical disease)
and case series and attack rate data. This example highlights the
need to use a range of techniques to fully understand the spatio-temporal
nature of disease occurrence. With knowledge of how disease occurs
in time and space, appropriate and effective disease control, prevention
and surveillance programmes can be implemented
Keywords
Clustering,
Disease, Epidemiology, Geographic information system, Space, Statistics,
Texas, Time, United States of America, West Nile virus.
|