Epidemic spreading in complex networks has received much attention in recent years. Previous research identified a propagation scenario of electronic threats which has not been described by any of the existing analytical models. In this scenario an infected node instead of being removed contributes to the infection spreading upon the reinfection attempt (for example, Sober, Sobig, and Mydoom Worms). In this paper we formally define and describe analytically a new model, Susceptible-Infected-Suspended-Reinfected (SISR), which complies with this scenario of epidemic spreading in both homogeneous and complex networks. We then evaluate the model by comparing it to the SIR model and by comparing its estimations with simulation results.