Spatio-temporal variation in the reported cases of waterborne infections obtained from surveillance systems

Elena N. Naumova
Tufts University School of Medicine

Objective. The goal of this study was to evaluate the utility of passive surveillance for waterborne infections in monitoring temporal and spatial variation in the incidence of these diseases and the relationship of that variation to the source of drinking water.

Methods. We abstracted all cases of giardiasis and cryptosporidiosis reported passively to the Massachusetts Department of Public Health for the period 1993-2000. Each reported case was linked to specially created databases indicating the source of residential water supply and socio-economic characteristics based on ZIP code of residence. We analyzed temporal and spatial patterns among reported cases using classical time series and spatial analysis.

Results. The geographic variations in these potentially waterborne diseases were associated with type of water supply, but may have been influenced by geographic variation in provider practice patterns regarding diagnosis and reporting. Passive surveillance data for giardiasis and cryptosporidiosis exhibited patterns of seasonal variations similar to those observed in active surveillance data from other sources. Seasonal trends also exhibited geographic variation in association with the category of water supply. Spatio-temporal analysis provides insights on potential sources of two cryptosporodiosis outbreaks in Worcester, 1995 and in Maynard, 1999.

Conclusions. Passive surveillance data collected consistently have the potential to provide valuable information on the temporal variation of disease incidence as well as geographical factors. However, passive surveillance data, particularly in the initial period of surveillance, may be highly sensitive to patterns of diagnosis and reporting and should be interpreted with caution.