Hand Foot and Mouth Disease has become an increasingly important emerging infection across South East Asia over the last decade. Highly episodic dynamics can result from periodic introductions of novel strains or from complex nonlinear dynamics. Disentangling these potential causes for observed dynamics requires sophisticated, computationally intensive model fitting methods along with detailed time series data on both infections and population processes such as birth and death rates. We analyze this question of strain introduction versus inherent dynamics (e.g. herd immunity) as the driving mechanism for observed pathogen dynamics using 12 years of weekly infection data on Hand Foot and Mouth Disease in Japan driven by the pathogen Enterovirus 71, a relative of poliovirus. We used a standard S-I-R (Susceptible-Infected-Recovered) model and maximum likelihood methods (iterated particle filtering) to test the hypothesis that herd immunity (inherent dynamics) rather than strain introduction drives outbreaks.
Results/Conclusions
Despite the irregularity of the data, and considerable observation error they reflect, we can discern a strong signature of nonlinear epidemic dynamics with complex coexisting attractors. Even with this complexity, the data are consistent with the pathogen acting like an immunizing (S-I-R) infection. This has implications for the potential success of future control measures (e.g. vaccination). We generalize our toolbox for disease ecologists, presenting a general platform to confront data with a range of epidemiological models to disentangle the dynamics of infectious disease, even in the case of severe, nonlinear dynamics.