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.

A bigger source of concern is if the increase in surveillance intensity focuses disproportionately on those cases that were exposed to the natural reservoir, as this would lead to overestimating F and underestimating R. For example, with the substantial increase in H3N2v-M virus infections during the summer of 2012, CDC changed their recommendations and asked clinicians to Figure 5 . Impact of uncertainty on the case detection rate and the overdispersion parameter on estimates of the reproduction number R. 12F always acts as a lower bound for R. Furthermore, an upper bound for R can be obtained if it is possible to specify an upper bound r max for the case detection rate and a lower bound k min for the overdispersion parameter k (see Text S1). The figure shows lower and upper bound for R as a function of r max . We specify k min~0 :16 which corresponds to the SARS scenario with superspreading events. doi:10.1371/journal.pmed.1001399.g005 Figure 6 . Estimating R and the case detection rate when both summary statistics F and G are available in surveillance scenario 1 (i.e., detection of a case does not trigger an outbreak investigation). In this simulation study, 10,000 clusters are generated for R = 0.5 and k = 0.5; six case detection rates are considered (1%, 10%, 20%, 30%, 40%, 50%; with a number of detected clusters that is 207, 1,706, 2,922, 4,092, 5,118, 6,125, respectively). First, the formula 12G gives point estimates for R (in the range 0.48-0.51 depending on the case detection rate). For given values of R and the overdispersion parameter k, it is possible to plot the relationship between F and the case detection rate. Black lines in the figure correspond to R = 0.5 (dashed line: k = 0.5; plain lines: k = 0.16 and k = 5). Colour triangles show estimates of the case detection rate obtained for each dataset when k is assumed to be known. When k is unknown, vertical colour plain lines give the range of values consistent with k in interval 0. 16 obtain respiratory specimens from ill persons with recent swine exposure [25, 26] . Therefore, in the summer 2012, ill persons with recent swine exposure may have been more likely to have been tested for H3N2v-M infection than those without such exposure. For this reason, we cannot use our method to analyze data collected in 2012.