Full length articleSocial network analysis and risk assessment: An example of introducing an exotic animal disease in Italy
Introduction
Exotic animal diseases (EADs) are transboundary hazards, characterized by their capability to cover global distances, affecting animal health and welfare with significant economic losses. Their prevention is complex and requires the dynamic management of potential entry points, transmission pathways and preventative barriers. The well-timed detection of an undefined or unexpected (exotic or re-emerging) threat could minimize the consequences due to onward transmission. Infectious diseases are transmitted among hosts by means of a variety of mechanisms, including direct contact, airborne and vector-borne transmission. The trade of live animals has been demonstrated to be an important mode of transmission for many diseases such as bovine tuberculosis, classical swine fever and Bluetongue (Fevre et al., 2006, Bigras-Poulin et al., 2006, Gilbert et al., 2005, Natale et al., 2009, Noremark et al., 2011). For this reason, livestock movements can be subject to strict controls to reduce the likelihood of disease transmission.
Risk analysis, in an animal health setting, is the formal process of assessing the risks associated with exposure to an imported infective agent (OIE, 2012). The first stage of this step-by-step process is the identification of the hazard, i.e. of the infectious pathogen able to cause adverse health effects in exposed populations; it is considered separately from the subsequent stage which is known as risk assessment. The latter is subdivided into three steps: (1) release (entry) assessment: description of the probable ways to introduce the hazard into a ‘free’ country; (2) exposure assessment: description of the routes necessary for exposure of animals to the hazard and the quantification of the characteristics and sizes of the exposed populations; (3) consequence assessment: quantification of the consequences, in terms of animal health, economic and environmental effects, occurring after the introduction and the establishment of the hazard (Peeler et al., 2015). The integration of the previous steps produces an overall estimation of the risk (OIE, 2012). A risk assessment model is a model focusing on the entrance of an exotic disease into a geographical area with naïve hosts. The risk assessment may also be performed in cases of sparse data but, in this case, will depend on many assumptions, as well as on subjective choices. For these reasons its success mostly lies in the choice of the most appropriate risk model (Nurminen et al., 1999).
One of the main aims of the collaborative European research project, SPARE, 'Spatial risk assessment framework for assessing exotic disease incursion and spread through Europe' [SPARE, 2016], was to develop a systematic methodology to rank the risk related to selected exotic livestock pathogens based on the probability of incursion and spread within Europe. An effective capability to recognise the emergence of a new or re-emerging disease relies upon a multiagency approach for the detection, diagnosis and surveillance of infectious disease as well as on an integrated methodological approach (King et al., 2004; Blancou et al., 2005, Walsh and Morgan, 2005). With this in mind, focusing on Bluetongue (BT) as an example, we integrated Social Network Analysis (SNA) into the classical exposure pathway used in SPARE with the aim of quantitatively determining the geographical risk of exposure of livestock to an exotic pathogen within Italy.
Social Network Analysis is a technique originally used to investigate the links among local patterns of social relationships within a social structure (Martinez-Lopez et al., 2009). In an epidemiological setting, the description of a social structure provides a flexible framework for investigating associations or interactions within a group. An advantage of the analytical approach of SNA, compared to other techniques, is the ability to handle bi-directional relationships within groups such as contacts among individuals, trade or animal movements. The elements of a network are: the nodes (or vertices or actors) and the connections among the nodes, referred to as edges (or contacts or links). SNA has different ways to be represented: (1) with a graph showing the set of pairs of nodes constituting the network; (2) with a list of the groups of elements and their interactions with a mathematical notation; (3) with the number of contacts represented by an adjacency matrix of NxN nodes, which are the number of contacts among the pairs of nodes in the network (Martinez-Lopez et al., 2009).
Bluetongue is an infectious viral disease transmitted by insect vectors and affecting ruminants. Sheep and goats are the most susceptible species and can show severe clinical symptoms, while in cattle infection is usually asymptomatic. This potentially presents an increased risk for spread of the virus via livestock movements as asymptomatic animals are less likely to be detected and removed from the system (https://www.gov.uk/government/news/bluetongue-virus-detected-and-dealt-with-in-imported-cattle). BT virus is almost exclusively transmitted by biting midges, such as Culicoides imicola, and the disease is generally considered to be present, or potentially present, in a zone between the 40° parallel North and the 35° parallel South, where climatic and environmental conditions are suitable to the vector's life cycle (Giovannini et al., 2004).
The aim of our work was to develop a spatially explicit risk assessment model to better estimate the spatial probability of the introduction of at least one BT affected animal by province within Italy per month. To better estimate spatial heterogeneity at a province level we integrated SNA into the standard risk assessment framework. The model was demonstrated using Italy as a case study but can be applied to any Member State (MS), provided the availability of appropriate animal movement data.
Section snippets
Overview
The model was based on a classical risk assessment pathway and was developed in R (v3.3.3, http://www.r-project.org/). The framework was modular, and each stage was developed independently with the results of the previous stage acting as the input for the next. The quantitative exposure assessment was stochastic with distributions chosen to best represent the uncertainty in the data.
The inputs for the model were provided by a release assessment developed by Simons et al., described in detail in
Results
The network built from the trade data is shown in the Fig. 4. This figure shows that majority of animals entering Italy go to provinces in the North West.
The map of the probability of the introduction of BT into Italy without taking into account the trade of livestock, i.e. for vector release only, is shown in Fig. 5. This suggests that the highest probability is in July and August in the southern provinces of Lecce and Taranto.
The median probability of having an autochthonous case represented
Discussion
Here, we have proposed a flexible model framework, adaptable to different animal diseases, for the spatial risk of an imported EAD which is useful to inform the preparedness of a Member State. The model was parameterised to assess the probability of onward transmission of BT virus, regardless of serotype, within individual Italian provinces. Our model estimated the probability of BT entering Italy with two different frameworks: one using a national probability, i.e. the same release probability
Acknowledgements
This work had funding agreed through the Animal Health and Welfare ERA-NET consortium (https://www.anihwa.eu/) under SPARE('Spatial risk assessment framework for assessing exotic disease incursion and spread through Europe'). Funders are acknowledged as the Department for the Environment, Food and Rural Affairs (Defra) – UK, Ministry of Health – Italy, Spanish National Institute of Agriculture and Food Research and Technology – Spain, and Federal Food Safety and Veterinary Office (FSVO) –
Conflict of interest
The authors have no conflict of interest to declare.
References (30)
- et al.
Assessing the impact of a cattle risk-based trading scheme on the movement of bovine tuberculosis infected animals in England and Wales
Prev. Vet. Med.
(2016) - et al.
Network analysis of Danish cattle industry trade patterns as an evaluation of risk potential for disease spread
Prev. Vet. Med.
(2006) - et al.
Animal movements and the spread of infectious diseases
Trends Microbiol.
(2006) - et al.
Mapping the basic reproduction number (R0) for vector-borne diseases: a case study on bluetongue virus
Epidemics
(2009) - et al.
Impact of human interventions on the spread of bluetongue virus serotype 8 during the 2006 epidemic in north-western Europe
Prev. Vet. Med.
(2008) - et al.
Network analysis of Italian cattle trade patterns and evaluation of risks for potential disease spread
Prev. Vet. Med.
(2009) - et al.
Network analysis of cattle and pig movements in Sweden: measures relevant for disease control and risk-based surveillance
Prev. Vet. Med.
(2011) - et al.
Emerging or re-emerging bacterial zoonoses: factors of emergence, surveillance and control
Vet. Res.
(2005) - et al.
Identification and review of sensitivity analysis methods
Risk Anal.
(2002) - et al.
Seasonal abundance and survivorship of Culicoides sonorensis (Diptera: ceratopogonidae) at a southern California dairy, with reference to potential bluetongue virus transmission and persistence
J. Med. Entomol.
(2000)
Cattle movements and bovine tuberculosis in Great Britain
Nature
The use of risk assessment to decide the control strategy for bluetongue in Italian ruminant populations
Risk Anal.
Modelling the effects of past and future climate on the risk of bluetongue emergence in Europe
J. R. Soc. Interface.
New partnerships between animal health services and public health agencies
Rev. Sci. Tech.
Social network analysis. Review of general concepts and use in preventive veterinary medicine
Transbound. Emerg. Dis.
Cited by (5)
Network analysis of pig movement data as an epidemiological tool: an Austrian case study
2023, Scientific ReportsSimulating contact networks for livestock disease epidemiology: a systematic review
2023, Journal of the Royal Society InterfaceApproaches for disease prioritization and decision-making in animal health, 2000–2021: a structured scoping review
2023, Frontiers in Veterinary ScienceA Link Analysis Based Approach to Predict Character Death in Game of Thrones
2020, EAI International Conference on Testbeds and Research Infrastructures for the Development of Networks and Communities (TRIDENTCOM)A link analysis based approach to predict character death in game of thrones
2020, Lecture Notes of the Institute for Computer Sciences, Social-Informatics and Telecommunications Engineering, LNICST