Effects of minimum epidemic and population sizes on a global epidemic in simulations of final size data

The stochastic SIR household epidemic model is well discussed in [2], [3] and [4]. The work of [1] • also proposed maximum likelihood based algori hm for its inference by assuming independence of epidemic in each household, contrary to the dependency assn uption in [4]. 8 ° * Using simulations, we examined the ne id for an appropriate choice of cut-off between small and large epidemics often referred to as minimum epidemic size, using rejection sampling, for a global infection to occur and then compared the estimates of the model parameters over a range of theoretical parameters, XL and Xc with corresponding z 6 [0,1]. We found that with large population size, appropriate choice of the minimum epidemic size and Xc =£ 0 facilitate the occurrence of a global epidemic. Thus, given these scenarios, the adequacy of the model fitness to the final size epidemic data is then realised.