Dealing with ecological fallacy in preparations for influenza pandemics: use of a flexible environment for adaptation of simulations to household structures in local contexts.

Development of strategies for mitigating the severity of a new influenza pandemic is a global public health priority. The aim of this study is to examine effects on simulation outcomes caused by variations in local socio-demographic data. We used a spatially explicit geo-physical model of a virtual city as a baseline and employed an ontology-modeling tool to construct alternative population distributions and household structures. We found that adjustment for the case when single parents in practice were cohabiting led to a higher reproduction rate than that observed for a population with the highest formally recorded share of households with >2 children. When antivirals sufficient to protect 10 percent of the population were administered to schoolchildren, a preliminary effect on the reproduction rate was observed. This effect was eliminated when the household structure was adjusted for cohabiting single parents. Nations have been encouraged to develop estimates of morbidity and mortality during a possible pandemic outbreak. In order to deal with ecological fallacy, the present results suggest that this recommendation can be extended also to local communities.