A representation in which individual hosts or epidemiological units are modelled as nodes and potential transmission interactions as edges (which may be weighted, directed, or time-varying), used to simulate how pathogen transmission depends on contact structure and to evaluate intervention strategies that exploit network heterogeneity.

Principle

Epidemic dynamics and control outcomes depend on network topology: node-level metrics (degree, centrality), clustering, community structure and temporal patterns modulate outbreak size, speed, and intervention efficiency.

Demonstration

Illustrative scenario: A livestock movement network records animal transfers between premises as directed edges. Simulating disease spread on this network shows that vaccinating or restricting movements at high-degree premises substantially reduces epidemic size compared with random vaccination, because these nodes act as hubs for onward transmission.

Misapplication

Using a static, aggregated contact network when contacts are highly time-dependent, or treating average contact rates as equivalent to the network structure; the semantic error is conflating mean-field mixing assumptions with explicit heterogeneity captured by network links.

Consequence

Enables targeted surveillance and control (e.g., targeted vaccination, movement restrictions, contact tracing) that exploit heterogeneity for efficiency; conversely, ignoring network structure can misallocate resources and underestimate superspreading potential.

Reversal

If contact patterns approximate homogeneous mixing (e.g., very high random mixing) or transmission is dominated by environmental reservoirs rather than direct contacts, network-based distinctions become less informative and compartmental models may suffice.

Boundary

Clearly within: models where transmission events are explicitly mapped to edges between identifiable units. Boundary case: coarse-grained networks that aggregate heterogeneous contacts into averaged links. Clearly outside: homogeneous compartmental models without explicit pairwise contact structure and purely environmental-transmission models without explicit contact edges.

Semantic Tension

Tension between the improved targeting and realism provided by detailed contact data and practical constraints of data privacy, completeness and temporal resolution; higher-fidelity networks improve inference but raise ethical and operational challenges.

Synthesis

Contact network models show that heterogeneity in who contacts whom is often the main driver of epidemic heterogeneity; efficient control therefore depends less on average rates and more on locating and altering the structure of high-risk connections.