Abstract

As a pathogen spreads throughout a human population network, understanding the time span between the first reported infection case and the establishment of local transmission relies on the ability to decompose infection incidence into local and travel cases, depending on whether the infected individual was exposed to the pathogen in their location of residence or elsewhere. However, most case data reported to public health agencies do not distinguish between local and travel-associated cases, hampering analysis of the critical early stages of the epidemic spread. We introduce an algorithm, based on the shape of the cumulative incidence curve, to estimate the time a pathogen takes to become locally established, based on the pathogen’s transmission and recovery rates and the network connectivity of the human population. This algorithm can predict the onset of an epidemic without considering any future case data, making it useful for tracking epidemics as they occur.