Addressing the problem of lack of representativeness on syndromic surveillance schemes

A major concern with some contagious diseases has recently led to an enormous effort to monitor population health status by several different means.
This work presents a modeling approach to overcome this poor data characteristic, allowing its use for the estimation of the true population disease picture. We use a state space model, where we run two processes in parallel - a process describing the non observable states of the population concerning the presence/absence of disease, and an observational process resulting from the monitoring.
We then use resampling importance sampling estimation techniques, in a Bayesian framework, which enables us to estimate the population states and, thus, the corresponding disease incidence curves.