Model hierarchies in edge-based compartmental modeling for infectious disease spread (Q376336)

The authors consider an edge-based compartmental modeling approach for the spread of SIR diseases in populations consisting of individuals with different contact dynamics. The populations are represented as networks, with nodes representing individuals. The nodes are then joined by edges representing potentially transmitting contacts. In particular, this approach incorporates the duration of contacts into the model, distinguishing between permanent, finite and fleeting contacts. The distribution of contact levels distinguishes between expected degree models and actual degree models. In expected degree models, each node is assigned an expected degree, the probability that an edge exists between two nodes being proportional to the expected degree of each node. In an actual degree model, each node is assigned an actual number of contacts, although not all of these contacts must be active at all times. After providing several considerations regarding the distinctions between expected degree and actual degree models, the authors indicate a flowchart that can be used to select an appropriate model from a given hierarchy of models (dormant contacts, dynamic variable-degree, dynamic fixed-degree, mean field social heterogeneity, mixed Poisson, configuration, mass action). In particular, after studying the convergence of models to the mass action model, the authors give conditions, under which the mass action model appropriately describes the spread of a disease.