El uso de Sistemas de Información Geográfica - CDIA-EC ...

El uso de Sistemas de Información Geografica y Sensores Remotos (SR) en Salud Pública
Bogotá, Colombia, 27 al 30 de Marzo de 2006
houses (e.g. Mott et al., 1978; Piesman et al., 1985; De Andrade et al., 1995;
Gurtler et al., 1998; Sosa-Jurado et al., 2004). In the 15 departments of Colombia
where the national survey was undertaken, it appears that incidence rates in
apparently infested houses are on average about 2%/year compared to 0.2% in
apparently uninfested houses, confirming that (1) there is some advantage to target
spraying to houses with detectable infestations, and (2) a risk map for triatomine
vectors should reflect the geographic distribution of current transmission rates.
Such a risk map, if suitably validated by comparison with independent data (eg
King et al 2004), could then be used to determine which strategy should be
implemented in each locality to make best use of limited resources.
Risk maps for triatomine vectors can be generated by a range of statistical
techniques, but they generally involve some combination of (1) interpolation from
the data collected in villages which have been surveyed and (2) statistical
associations with environmental features – such as temperature, rainfall, land
cover, and socioeconomic variables. To date all risk maps that have been
described for Chagas disease vectors have been static (Gorla 2002, Costa et al
2002, Peterson et al 2002, Beard et al 2003, Dumonteil & Gorbiere 2004, Lopez-
Cardenas et al 2005). Whereas a decision-support system for Chagas disease
control programmes would benefit most from a dynamic risk map, where
predictions of risk are allowed to vary each year according to changing conditions,
most notably in response to house spraying. Such a dynamic map could allow re-
infestation rates of sprayed villages to vary geographically in response to variability
in local environmental conditions (whether vegetation- or household-related).
Evaluations of the effectiveness of house spraying against T. dimidiata infestations
in Guatemala, for example, indicate some geographic clustering of locations where
spraying has been less effective (Hashimoto K. & Codron-Rosales C., pers.
comm.).
In conclusion, it is worth emphasising that the introduction of Geographic
Information Systems into Chagas disease control programmes should not only aid
in the development of evidence-based control strategies (i.e. policy), but will also
provide immediate help by improving the quality of data management and so
improve operational capability (i.e. practice). There is considerable scope to make
27