MODELING THE DYNAMICS OF ENTEROVIRUS-71 IN TAIWAN: An application of the TSIR model

Abstract

Background: Human Enterovirus-71 (EV-71) is an emerging pathogen that has become a novel target for disease surveillance and control efforts in the Asia-Pacific region. EV- 71 and the group A coxsackieviruses are the major causative agents of Hand, Foot & Mouth Disease (HFMD), normally a self-limiting childhood infection transmitted via the fecal-oral route that causes a rash of vesicular lesions on the hands, feet and oral mucosa. The majority of individuals infected with HFMD require little to no medical attention; in some cases, however, enterovirus infection leads to severe neurologic complications like encephalitis, meningitis, myocarditis and acute paralysis. Research goals: EV-71 poses a disease modeling challenge, as existing data captures only severe cases and not the true burden of infection. This thesis aims to mathematically derive the true burden of EV-71 infection in Taiwan, to elucidate key epidemiological parameters from available data using a stochastic, discrete-time modeling approach, and to use those parameters to model the potential impact of vaccination. Methods: Data were sourced from publicly available databases maintained by the Taiwan Center for Disease Control. The Time series-Susceptible-Infected-Recovered (TSIR) model was used for parameter inference and simulation. Results: Key modeling parameters were elucidated from the dataset, including 53 weekly transmission parameters, an overall contact mixing parameter and a reconstructed susceptibles class. The true burden of infection was estimated to be 1492 times the reported number of cases. Simulations forecasting vaccination identified a critical population coverage level at 99.95%. Conclusion: In the case of severe enterovirus infection, the TSIR modeling approach provides a useful tool for the extrapolation of epidemiological parameters from a limited dataset. If proper forecasting is to be done for EV-71, however, more work