Search

African swine fever (ASF) is considered the most impactful transboundary swine disease. In the absence of effective vaccines, control strategies are heavily dependent on mass depopulation and shipment restrictions. Here, we developed a nested multiscale model for the transmission of ASF, combining a spatially explicit network model of animal shipments with a deterministic compartmental model for the dynamics of two ASF strains within 3 km × 3 km pixels in one Brazilian state. The model outcomes are epidemic duration, number of secondary infected farms and pigs, and distance of ASF spread. The model also shows the spatial distribution of ASF epidemics. We analyzed quarantine‐based control interventions in the context of mortality trigger thresholds for the deployment of control strategies. The mean epidemic duration of a moderately virulent strain was 11.2 days, assuming the first infection is detected (best‐case scenario), and 15.9 days when detection is triggered at 10% mortality. For a highly virulent strain, the epidemic duration was 6.5 days and 13.1 days, respectively. The distance from the source to infected locations and the spatial distribution was not dependent on strain virulence. Under the best‐case scenario, we projected an average number of infected farms of 23.77 farms and 18.8 farms for the moderate and highly virulent strains, respectively. At 10% mortality‐trigger, the predicted number of infected farms was on average 46.27 farms and 42.96 farms, respectively. We also demonstrated that the establishment of ring quarantine zones regardless of size (i.e. 5 km, 15 km) was outperformed by backward animal movement tracking. The proposed modelling framework provides an evaluation of ASF epidemic potential, providing a ranking of quarantine‐based control strategies that could assist animal health authorities in planning the national preparedness and response plan. [ABSTRACT FROM AUTHOR]

Tracking animal movements over time may fundamentally determine the success of disease control interventions. In commercial pig production growth stages determine animal transportation schedule, thus it generates time‐varying contact networks showed to influence the dynamics of disease spread. In this study, we reconstructed pig networks of one Brazilian state from 2017 to 2018, comprising 351,519 movements and 48 million transported pigs. The static networks view did not capture time‐respecting movement pathways. For this reason, we propose a time‐dependent network approach. A susceptible‐infected model was used to spread an epidemic over the pig network globally through the temporal between‐farm networks, and locally by a stochastic model to account for within‐farm dynamics. We propagated disease to calculate the cumulative contacts as a proxy of epidemic sizes and evaluate the impact of network‐based disease control strategies in the absence of other intervention alternatives. The results show that targeting 1,000 farms ranked by degree would be sufficient and feasible to diminish disease spread considerably. Our modelling results indicated that independently from where initial infections were seeded (i.e. independent, commercial farms), the epidemic sizes and the number of farms needed to be targeted to effectively control disease spread were quite similar; indeed, this finding can be explained by the presence of contact among all pig operation types The proposed strategy limited the transmission the total number of secondarily infected farms to 29, over two simulated years. The identified 1,000 farms would benefit from enhanced biosecurity plans and improved targeted surveillance. Overall, the modelling framework provides a parsimonious solution for targeted disease surveillance when temporal movement data are available. [ABSTRACT FROM AUTHOR]