.-
help for ^outbrk^                                                 (STB-54: sbe32)
.-

Automated outbreak detection from public health surveillance data
-----------------------------------------------------------------

    ^outbrk^ #week #year

where

     #week is the number of the current week

     #year is the number of the current year


Description
-----------

^outbrk^ calculates threshold values for outbreak detection of infectious
diseases. The data are weekly reports of positive microbiological diagnostics
from a passive surveillance system based on laboratory reporting.

The procedure is based on a log-linear regression model, adjusted for
overdispersion, seasonality, secular trends and past outbreaks. This model is
used to calculate an expected value for the current week based on historical
data, together with a warning threshold value above which an observed count is
declared to be unusual.

When no cases are reported in a week, we asume that no outbreak occurred and
thus no model is fitted. As a consecuence, no threshold value is calculated.

The output is a table displaying the list of microorganisms with the observed
number of cases and the threshold value for the current week. In addition, a
warning message is displayed when the actual report exceed the threshold.

^outbrk^ works after setting the variables with the ^varset^ ado 


Example
-------

        . ^outbrk  3 1998^


Authors
------

        Esther Lopez, Isolina Santiago, Luisa Abraira.
        Direccion Xeral de Saude Publica, Xunta de Galicia.
        dxsp3@@jet.es


Also see
--------

On-line:  help for @varset@