Surveillance of influenza from household to community in Taiwan

Abstract

We propose a chain binomial model to elucidate the transmission modes of influenza within households and with which to further evaluate the spread of influenza in a community. Data on the surveillance of influenza in Taiwan were collected from July 2001 to March 2002. A series of chain binomial models were fitted to estimate the escape probability. This parameter together with accumulated cases of the first episode were further applied to derive the transmission probability, the contact rate and the recovery rate, yielding an estimate of the basic reproductive number and providing a simulation approach with which to assess the spread of influenza in the community. Of the 6 391 614 cases identified during the study period, 4 451 046 were from 3 654 320 households consisting of two to five people, with at least one case used to estimate the escape probabilities within households. The escape probability within households was 92.4%, which, in turn, led to a contact rate estimate of 7.7 per day. The transmission probability was estimated as 8.3%. These three parameters yielded 2.56 as the basic reproductive number. The proposed method provides a means for using household data to evaluate the transmissibility of influenza within a community.