Using Routine Surveillance Data to Estimate the Epidemic Potential of Emerging Zoonoses: Application to the Emergence of US Swine Origin Influenza A H3N2v Virus

Cauchemez, Simon; Epperson, Scott; Biggerstaff, Matthew; Swerdlow, David; Finelli, Lyn; Ferguson, Neil M.

Our approach can also be used to evaluate risks associated with non-zoonotic disease. As an example, we consider the cholera outbreak that started in the Dominican Republic in 2010 following the epidemic of that disease in Haiti. We can assess the level of Figure 2 . Probability F that the first detected case in a cluster was infected by the reservoir, as a function of the reproduction number R. The number of secondary cases caused by individual human cases is modelled with a Negative binomial distribution with parameters R and k, where k is the overdispersion parameter [12] . We consider three scenarios for the case-to-case variation in infectiousness: high (i.e., most transmission events are caused by a small proportion of cases like for SARS; k = 0.16), medium (k = 0.5), and low (k = 5) [12] . doi:10.1371/journal.pmed.1001399.g002 local transmission in the Dominican Republic from the proportion G of cases detected in the Dominican Republic that were linked to Haiti (the latter country here playing the role of the natural reservoir). By December 18, 2010, three of the 59 confirmed cases detected in the Dominican Republic were linked to Haiti (G = 5%) [22] . Our simple estimation method cannot distinguish between R<1 and R.1; but the observation that G was close to zero is indicative that transmission in the Dominican Republic was close or above levels needed for sustained human-to-human transmission.