Estimating the extent of household contact misclassification with index cases of disease in longitudinal studies using a stochastic simulation model