Why Integral Equations should be used instead of differential equations to describe the dynamics of epidemics