AEGISS Data on the Spatio-Temporal Distribution
of Non-specific Gastrointestinal Disease in
Hampshire, UK

These data may be used by bonafide researchers for non-commercial
purposes. They are offered in good faith, but I cannot guarantee 
their accuracy. If you use the data in published work, please
cite the reference below and include an acknowledgement along the 
following lines.
 
ACKNOWLEDGEMENT. Data were provided by Prof.Peter Diggle, Lancaster
University and Dr Peter Hawtin, Health Protection Agency,Southampton Laboratory.
The data were derived from anonymised data-sets collected during
project AEGISS, a collaborative surveillance project based in the south
of England.  The project was financially supported by the Food Standards Agency.

CITATION: Diggle, P, Knorr-Held, L, Rowlingson, B, Su, T, Hawtin, P
and Bryant, T (2003). On-line monitoring of public health surveillance 
data. In "Monitoring the Health of Populations: Statistical Principles
and Methods for Public Health Surveillance" editors R. Brookmeyer and
D.F. Stroup, pages 233-66. Oxford : Oxford University Press.

DESCRIPTION:

File AEGISS_ixyt.dat contains id numbers and space-time locations,
(id,x,y,t) for 10,572 cases of non-specific gastrointestinal 
disease in the county of Hampshire, UK, as reported to NHS Direct
(a phone-in triage service operating within the UK's National
Health Service). Each (x,y)-location corresponds to the centroid of 
the unit post-code of the residential address of the person making 
the call to NHS Direct. The unit of distance is 1 metre. The unit of 
time is 1 day, with day 1 corresponding to 1 January 2001.

File AEGISS_poly.dat contains vertices of a 120-sided polygon
representing the boundary of the study-region (Hampshire,UK).
The unit of distance is 1 kilometre.

Using the splancs package within R, a picture of the spatial locations 
can be constructed as follows:

library(splancs)
data<-read.table("AEGISS_ixyt.dat",header=T)
xy<-cbind(data$x,data$y)
poly<-1000*matrix(scan("AEGISS_poly.dat"),120,2,T)
par(pty="s")
polymap(poly)
pointmap(xy,add=T,pch=".")

                                     Peter Diggle, 25 September 2005