An epidemic model to evaluate the homogeneous mixing assumption

• Epidemic models can be described by ordinary differential equations.
• In these models, heterogeneities concerning the host contact network are neglected.
• We propose a model to evaluate the validity of this homogeneous mixing assumption.
• We perform this evaluation by using real epidemic data.
• We also show that the model can experience transcritical and Hopf bifurcations.

Many epidemic models are written in terms of ordinary differential equations (ODE). This approach relies on the homogeneous mixing assumption; that is, the topological structure of the contact network established by the individuals of the host population is not relevant to predict the spread of a pathogen in this population. Here, we propose an epidemic model based on ODE to study the propagation of contagious diseases conferring no immunity. The state variables of this model are the percentages of susceptible individuals, infectious individuals and empty space. We show that this dynamical system can experience transcritical and Hopf bifurcations. Then, we employ this model to evaluate the validity of the homogeneous mixing assumption by using real data related to the transmission of gonorrhea, hepatitis C virus, human immunodeficiency virus, and obesity.