Network analysis of pig movements in Ecuador: Strengthening surveillance of classical swine fever

Abstract

The analysis of domestic pig movements has become useful to understand the disease spread patterns and epidemiology, which facilitates the development of more effective animal diseases control strategies. The aim of this work was to analyse the static and spatial characteristics of the pig network, to identify its trading communities and to study the contribution of the network to the transmission of classical swine fever. In this regard, we used the pig movement records from the National Veterinary Service of Ecuador (2017-2019), using social network analysis and spatial analysis to construct a network with registered premises as nodes and their movements as edges. Furthermore, we also created a network of parishes as its nodes by aggregating their premises movements as edges. The annual network metrics showed an average diameter of 20.33, a number of neighbours of 2.61, a shortest path length of 4.39 and a clustering coefficient of 0.38 (small-world structure). The most frequent movements were to or from markets (55%). Backyard producers made up 89% of the network premises, and the top 2% of parishes (highest degree) contributed to 50% of the movements. The highest frequencies of movements between parishes were in the centre of the country, while the highest frequency of movements to abattoirs was in the south-west. Finally, the pattern of classical swine fever (CSF) disease outbreaks within the Ecuador network was likely the result of network transmission processes. In conclusion, our results represented the first exploratory analysis of domestic pig movements at premise and parish levels. The surveillance system could consider these results to improve its procedures and update the disease control and management policy, and allow the implementation of targeted or risk-based surveillance.

Risk Factors and Spatiotemporal Analysis of Classical Swine Fever in Ecuador

“…More than half of the case notifications (51%) were reported by the owner, followed by health sensors (42%), which are volunteers selected by the NVS to enhance the surveillance system directly from communities across the country. There was a higher proportion of cases in the third community network, known for their high density of backyard producers and indigenous communities [ 47 ], located in the centre of the country. The highest proportion of cases number was in commercial production ( Table 3 ).…”

“…Some cases occurred on premises with previous vaccination records, possibly related to illegal movements of unvaccinated animals [ 47 ] or vaccination failures because of poor handling and mismanagement, as evidenced in neighbouring Colombia [ 4 , 20 ]. The risk associated with higher temperatures in the spatiotemporal analysis could also be linked to problems in the vaccine cold-chain in regions such as the Amazon and the coast, which represent a logistical challenge when addressing average temperatures above 20 degrees Celsius, corroborating that vaccination strategies alone may not eradicate the disease [ 58 , 59 ]; Furthermore, possible antigenic alterations because of vaccination pressure and their effects on the epidemiology of the disease could also be further analysed [ 60 , 61 ].…”

Uma contribuição para a avaliação do risco de entrada de pragas exóticas via comércio internacional no Brasil

“…Sendo assim, o modelo já foi utilizado por diversos autores para estudar a disseminação de doenças em humanos (CHRISTLEY et al, 2005), animais (MARTÍNEZ-LOPES et al, 2009;MENEZES et al, 2020;ACOSTA et al, 2022) e plantas (LEFEBVRE et al, 2011; PAUTASSO e JEGER, 2014). Lefebvre et al (2011) discutem como os conceitos de redes podem ser aplicados na fitopatologia uma vez que as conexões possuem papel fundamental em como uma epidemia se espalha.…”