In public health emergencies, infectious disease transmission often varies significantly across populations. To investigate population heterogeneity and interaction mechanisms in cross-population transmission dynamics, we develop a community-based SIR (susceptible-infectious-removed) compartmental model applicable to multiple populations. This model is grounded in a heterogeneous network that incorporates both intra-population structural differences and inter-population coupling. Considering the structural characteristics and interactions between the service-providing and the served populations, we introduce multidimensional parameters, including contact preferences, different transmission rates, and recovery rates. The basic reproduction number is obtained analytically through the next-generation matrix method, with the stability of disease-free and endemic equilibria rigorously proved. Agent-based simulations and multi-scenario experiments are conducted to examine how network structure, coupling strength, and population migration mechanisms shape transmission dynamics. Finally, early-stage COVID-19 epidemic data from Hubei Province are used for parameter fitting, and the fitted parameters are employed to validate the Monte Carlo and Mean-field models, demonstrating the high accuracy of the model. This study shows that the coupling of the service population significantly accelerates epidemic transmission. Furthermore, population migration is identified as a transmission accelerator that exhibits a non-monotonic impact on the steady-state infection scale. However, under resource-constrained conditions, prioritizing the improvement of the recovery rate of small-scale service populations can indirectly reduce the final epidemic size in large-scale community populations. This study offers quantitative guidance for designing targeted interventions under multidimensional heterogeneity.
Zhu et al. (Sun,) studied this question.