key: cord-283316-a8jewy2h authors: Bianchini, Juana; Humblet, Marie‐France; Cargnel, Mickaël; Van der Stede, Yves; Koenen, Frank; de Clercq, Kris; Saegerman, Claude title: Prioritization of livestock transboundary diseases in Belgium using a multicriteria decision analysis tool based on drivers of emergence date: 2019-10-09 journal: Transbound Emerg Dis DOI: 10.1111/tbed.13356 sha: doc_id: 283316 cord_uid: a8jewy2h During the past decade, livestock diseases have (re‐)emerged in areas where they had been previously eradicated or never been recorded before. Drivers (i.e. factors of (re‐)emergence) have been identified. Livestock diseases spread irrespective of borders, and therefore, reliable methods are required to help decision‐makers to identify potential threats and try stopping their (re‐)emergence. Ranking methods and multicriteria approaches are cost‐effective tools for such purpose and were applied to prioritize a list of selected diseases (N = 29 including 6 zoonoses) based on the opinion of 62 experts in accordance with 50 drivers‐related criteria. Diseases appearing in the upper ranking were porcine epidemic diarrhoea, foot‐and‐mouth disease, low pathogenic avian influenza, African horse sickness and highly pathogenic avian influenza. The tool proposed uses a multicriteria decision analysis approach to prioritize pathogens according to drivers and can be applied to other countries or diseases. the OIE from different European member states including Belgium (World Organisation for Animal Health, 2018) . Another very important recent emerging livestock disease reported specifically in Belgium at the end of 2018 was African swine fever, although cases so far have been reported only in wild boars . Its emergence is of great concern for the pig industry of the region and being a disease, which until now has been exotic for Belgium. It shows how diseases may re-emerge unexpectedly with most likely origin attributable to human activity (Saegerman, 2018) . The (re-)emergence of diseases shifts in relation to several underlying set of factors inherent to modern society, that is the so-called 'drivers'. The joint presence of these drivers can create an environment in which infectious disease can (re-)emerge and be maintained in animal and/or human compartments (King, 2004) . Many drivers have been identified, such as climate change, global travel, immigration patterns, increase in the human population, environmental degradation and others (Altizer, Ostfeld, Johnson, Kutz, & Harvell, 2013; Daszak, Cunningham, & Hyatt, 2000; King, 2004) . The threat of (re-)emergence is more likely to increase and past experience has shown that no country, however economically welldeveloped it may be, is capable of ensuring 100% security of its borders, even by imposing measures such as quarantine protocols or import bans on animals and animal products (Ben Jebara, 2004) . In Belgium, the monitoring and reporting of livestock diseases are subjected mostly on self-reporting of suspected clinical cases by the farmers to the Federal Agency for the Safety of the Food Chain (FASFC), with an established list of mandatory notifiable diseases for livestock and other species (aquatic, exotic) (Federal Agency for the Safety of the Food Chain, 2019). Each suspicion is then confirmed by laboratory analysis (Federal Agency for the Safety of the Food Chain, 2019). Thus, a rational priority setting approach is needed to assist decision-makers in identifying and prioritize diseases that are more likely to (re-)emerge and as such allocating the right resources tailored to a particular disease threat. One such approach used is disease prioritization, which has as main objectives: to optimize financial and human resources for the surveillance, prevention, control and eradication of infectious disease and to target surveillance for early detection of any emerging diseases (Humblet et al., 2012) . Some studies identified key characteristics of potential emerging infectious diseases and prioritized infectious diseases according to their risk of (re-)emergence or impact in some countries (Cardoen et al., 2009; Cox, Sanchez, & Revie, 2013; Havelaar et al., 2010; Humblet et al., 2012) . Hence, these focused on human or zoonotic diseases and the impact they would have in certain countries. In this study, the focus is livestock epidemic diseases and the aim was to identify (re-)emergence drivers' criteria and with it use expert elicitation to prioritize livestock epidemic diseases that may emerge in Belgium. A multicriteria decision analysis (MCDA) method was chosen because it provides a systematic way to integrate information from a range of sources (Cox et al., 2013) and it aims to improve transparency and repeatability (European Centre & for Disease Prevention & Control, 2015) . Multicriteria decision analysis requires identifying criteria and scoring criteria according to the pathogen/disease. By weighting each criterion and calculating weighted scores from the criteria, an overall score per pathogen/disease was calculated (European Centre & for Disease Prevention & Control, 2015; Humblet et al., 2012) . This is the first study to prioritize livestock epidemic disease using drivers as criteria. This prioritization list could be an aid to decision-makers to make an informed decision on course of actions to be taken and use the correct resources when there is a threat of a disease (re-)emerging in Belgium. We compiled a list of livestock-associated infectious diseases ( Figure 1) using a systematic approach. This was done by collating in a single database notifiable terrestrial animal diseases from different governmental official lists from Belgium (Federal Agency for the Safety of the Food Chain, 2015) and neighbouring countries (Luxembourg was excluded because of high similarity), that is Germany (Federal Ministry of Food & Agriculture of Germany, 2015), France (Légifrance, 2015a (Légifrance, , 2015b , the Netherlands (Ministerie van Landbouw, 2015) and Great Britain (Scottish Government, 2015) . In order to broaden the spectrum, diseases included in two other lists of official international organizations, that is the World Organisation for Animal Health (OIE) (World Organisation for Animal Health, 2015) and the European Union (European Commission, 2012) , were also added to the database. Only diseases that affect cattle, sheep, goats, swine and poultry (livestock) were selected from the official lists and included in database. (Azkur et al., 2013; Chaintoutis et al., 2014; Lievaart-Peterson et al., 2012; Yilmaz et al., 2014) . Thus, the risk of any of these viruses to (re-) emerge may be present, which further prompted the necessity of adding these three viruses to the list of diseases to be prioritized. The main objective was to prioritize the diseases according to drivers of (re-)emergence. A driver was defined as a factor, which has the potential to directly or indirectly precipitate ('drive') or lead to the (re-)emergence of a livestock infectious disease. We identified different criteria considered as drivers through scientific literature and previous disease prioritization exercises, and discussion with experts from academia, government agencies and international bodies. A total of 50 criteria were identified and classified under 8 different domains (Table 1) Each criterion had a definition of the coefficient, which ranged from 0 to 4 accordingly (Appendix A). Each domain spreadsheet had a number of criteria. For each criterion, coefficients were clearly defined for a good comprehension and standardization. Coefficients were from scores of 0 to 4 or from 1 to 4 (a number of criteria could not be scored with a zero, e.g. current F I G U R E 1 Systematic process for selecting the livestock diseases. * Livestock diseases were those which affected cattle, sheep, goats, swine and poultry TA B L E 1 List of criteria used to prioritise (re)emerging infectious diseases, according to their likelihood of (re)emergence in Belgium in response to different categories of drivers (Gore, 1987) . The number of points to be distributed was proportional to the number of criteria per category multiplied by ten. Indeed, the criterion with the most points allocated is considered the one that weighs the most in the category. If, on the other hand, all the criteria have the same weight in the category, the distribution is equitable, with 10 points for each criterion. For example, 90 points were to be distributed between the 9 criteria of the 'pathogen characteristics' domain. Indeed, the criterion with the most points allocated is considered the one that weighs the most in the pathogen characteristics. Such process illustrated the experts' opinion on the relative importance of criteria within one domain. The last spreadsheet was dedicated to the inter-domain weighting. Experts were asked to distribute a total of 80 points (N = 8 domains) among the domains to classify the domains according to their opinion. Two rounds of expert elicitation were implemented. The first round consisted in the questionnaire assessment; experts were asked to verify whether the questions were in relation with the drivers and whether the scoring systems were correctly defined and identified. The questionnaire and related instructions were sent to 14 experts (Appendix B) by e-mail. The experts were asked to complete questionnaire by scoring and additionally to assess and give comments on the criteria and coefficient definitions. The questionnaire was then refined according to experts' comments and suggestions. For the second round, 62 experts were identified (Appendix C) via Internet searching and recommendations from the project partners and recruited participants. These experts were asked to answer the questionnaire in order to rank the diseases. Thus, they had to choose the defined coefficient for each criterion (i.e. criterion scoring), then distribute the points for within each domain (i.e. the intradomain weighting), and lastly distribute the points within the domains (i.e. inter-domain weighting). They were invited to participate via a project summary e-mail and were sent the reviewed questionnaire via e-mail if they agreed to participate. Experts were recruited until a minimum of 4 experts per disease was obtained with a maximum of 5 experts. In some cases, one expert could answer several questionnaires (one per disease) if the diseases were within is area of expertise. To obtain the overall score for the ranking, an aggregation method that combined the 2 types of weighting (i.e. the intra-and inter-domain) was used. First, the criterion score (coefficients attributed by experts) had to be standardized. Indeed, some criteria were allocated coefficients from 0 to 4 and others from 1 to 4. This standardized score was then multiplied by the intradomain weight as given by the expert. These results were summed to obtain a domain score. In this formula, DSj = domain score, crit = criterion, SCj = standardized score of the criterion and WdWj = intradomain weight for each criterion. Each domain score was then multiplied by the inter-domain weight. These results were summed and an overall weighted score calculated, per expert and per disease. In this formula, OWS = overall weighting score of each expert for a specific disease, cat = category, DSj = domain score and IdWj = inter-domain weight. Each disease had 4 or 5 OWS (since there were 4 or 5 experts per disease), and thus, for each disease, the final score was the average of all disease experts' OWS. The final score was then used to rank the diseases, based on drivers, from the highest score to the lowest. The highest score corresponded to the disease with the highest risk of (re-) emerging according to the drivers. In addition, the median and range among the scores of all the disease experts were also obtained. With the median, a ranking was done to observe whether there was any significant difference with the ranking obtained using the mean. The range was used to note which diseases had the highest and lowest level of variation/uncertainty among the final experts' average score. In order to determine which driver(s) was/were considered as the most influential for the (re-) emergence of diseases, the domains were ranked. Domain ranking was performed using the inter-domain scores (weights). The sum of each domain weight (∑IdWj) per disease and per domain given by each expert was ranked from the high to the low, that is 1 to 8. Then, for each domain, the frequency of their rank was used to display in graph. A cluster analysis was implemented using regression tree analysis (Salford Predictive Modeler ® , Version 8.2, Salford Systems, San Diego, California, USA). The normalized disease score is a continuous variable, and the aim was to obtain groups in qualitative categories of importance (e.g. very high, high, moderate and low) with minimal within-group variance. Two sensitivity analyses were assessed, that is on expert elicitation and influence of a domain. This was achieved by repeating the disease ranking with a 'reduced' version of the model and comparing the new ranking to the complete model. The experts' sensitivity analysis consisted in dividing them into 4 groups. Scores were then re-calculated by deleting a group of experts. Each reduced ranking model was compared to the full complete model by using the Spearman's rank test to establish whether the ranking was correlated between the complete and the reduced models. The sensitivity analysis on the domains was done by deleting one domain and re-calculating the mean scores to rank the diseases. This 'reduced' ranking was then compared with the complete model, and the Spearman's rank test was applied. If the ranking position changed to less than three places, then the final score was considered as robust. If it changed to more than two places, then it was considered as a domain of drivers influencing greatly disease (re-) emergence. We compiled a list of 29 diseases ( All 14 experts contacted for the first phase (questionnaire assessment) answered positively (Appendix B). There was a general agreement on which criteria and coefficients were clear or not. Neither criterion nor coefficient were deleted but only amended according to experts' suggestions. For the second phase of expert elicitation, a total of 62 experts agreed to participate and answered the questionnaires (Appendix C). The objective of minimum of 4 experts per disease was reached, and the maximum of 5 experts was reached for 8 diseases. The final disease ranking based on the average final scores is shown in Figure 2 . The higher the mean score, the higher the ranking, which means the disease is most likely to (re-)emerge in Belgium. The top 5 diseases in decreasing order were porcine epidemic diarrhoea (PED), FMD, low pathogenic avian influenza (LPAI), African horse sickness (AHS) and HPAI (Table 3) . On the other end, the diseases with the lowest mean scores were haemorrhagic septicaemia, Japanese encephalitis, WNF, peste des petits ruminants (PPR) and Nipah disease. (1) The regression tree analysis determined 4 clusters ( Figure 2 ). The clusters distinguished five, eleven, nine and four diseases, and were classified, respectively, as of 'low importance', 'moderate importance', 'high importance' and 'very high importance' (i.e. highly influenced by drivers). The diseases belonging to the node 'highest importance' were PED, FMD, LPAI and AHS. The node of the lowest importance included haemorrhagic septicaemia, Japanese encephalitis, PPR, Nipah disease and WNF. The relative importance of the 8 domains varied depending on the disease. However, when considering all domains for all 29 diseases, 'economy and trade activities' obtained the highest number of points, being ranked first 15 times and zero times last ranked (8th). The opposite can be said about 'characteristics of farm/production system', as it was never ranked 1st nor 2nd ( Figure 3 ). The sensitivity analysis done on the groups of experts showed that the ranking of diseases was not affected in the reduced models. Indeed, the Spearman's rank-order correlation indicated a strong positive association of ranks when using different groups of experts for different reduced models, showing that there was a consistency among the scoring of the experts. TA B L E 2 (Continued) As for the domain sensitivity analysis, Table 3 The MCDA approach allowed the selection of 29 livestock diseases exotic to Belgium and their prioritization based on drivers. Whilst such an approach was used in previous disease prioritization exercises, this is one of the first to consider livestock epidemic diseases only and to use criteria related to drivers of (re-)emergence. Only diseases exotic to Belgium were prioritized. The diseases that fitted the eligibility criteria were all of viral origin, except haemorrhagic septicaemia (Pasteurella multocida, serotypes 6:B, 6:E), CCPP and CBPP. Few zoonoses were included in the list (n = 6) as the prioritization exercise focused on livestock epidemic diseases. Therefore, several zoonoses included in other prioritization processes were excluded. Regarding prioritization, PED ranked top of the list. Although currently not reportable neither in the EU (except in the UK) nor to the OIE, it ranked high in all models (high mean score), possibly due to its highly transmissible character and the difficulty to control it; furthermore, the disease mainly concerns intensive production. Cases have already been reported in EU Member States: for example in May 2014, an outbreak of diarrhoea occurred in fattening pigs on German farms. An outbreak of diarrhoea occurred on a Belgian fattening pig farm at the end of January 2015; this was the first confirmed PED case in Belgium in decades (Theuns et al., 2015) . When the list of diseases was compiled, the outbreak had not occurred yet, but when the experts answered the questionnaire it had, and therefore, this was most likely the reason why it ranked at the top of the prioritized list. Low pathogenic avian influenza ranked slightly higher than HPAI in this multicriteria analysis on the risk of (re)-emergence (LPAI ranked 3rd whilst HPAI ranked 5th (1) (2) (2) (Monne et al., 2014) . Hence, these characteristics of the virus give in this prioritization LPAI a higher score than HPAI, but HPAI is more likely to be detected and notified. African horse sickness surprisingly ranked 4th, although its last know incursion in Europe (Portugal and Spain) was in 1987 and its eradication dates back to 1990. Such high position in the ranking could be related to its vector-borne transmission, that is by Culicoides biting midges. These vectors are often highly abundant, across most of Africa, the Middle East, Europe and southern Asia (Carpenter, Mellor, Fall, Garros, & Venter, 2017) . Additionally, the recent changes in the epidemiology of bluetongue and its latest epidemic in Europe and the emergence of Schmallenberg disease (Afonso et al., 2014; Anonimous, 2013; Carpenter et al., 2009; highlight the uncertainty about the variables controlling the spread and persistence of Culicoides-borne arboviruses. These different factors have raised concerns that AHS may also amount similar incursions, hence explaining such high mean final score in the prioritization process. In this prioritization, most of the diseases were in clusters 2 (high importance, N = 9) and 3 (moderate importance, N = 11 This score can only be compared with the prioritization work done by Humblet and collaborators (Humblet et al., 2012) as other prioritization works using the MCDA method, such as those by Cardoen et al. (2009), and Havelaar et al. (2010) , only included zoonoses. Indeed, in regression tree analysis of prioritized diseases of food-producing animals and zoonoses, ASF also fell in the 3th group of importance out of the 4th group (Humblet et al., 2012) , just like in this prioritization work. Another study, which may be used for comparison as it used MCDA approach and had swine diseases, done by Brookes, Hernandez-Jover, Cowled, Holyoake, and Ward (2014), ASF ranked higher, but in this study only exotic diseases for the pig industry in Australia were ranked using criteria related to impact and the experts were pig producers, which changes the importance in the scores, giving ASF a higher ranking. The livestock diseases at the bottom of the list were Nipah disease, PPR, WNF, Japanese encephalitis and haemorrhagic septicaemia. In other prioritization exercises, Nipah, Japanese encephalitis and WNF were ranked in a higher category (Cox et al., 2013; Havelaar et al., 2010; Humblet et al., 2012) . The prioritiza- Drivers are a complex set of factors, and their convergence can cause the (re-)emergence of a disease. Several drivers have a stronger impact on diseases compared to others, as shown in the results section. Porcine epidemic diarrhoea ranked at the top in all models, except in the reduced models of production system characteristics. Porcine epidemic diarrhoea affects mainly intensive production systems; thus, the driver category 'production system characteristics' logically influences a lot. When using the reduced model, the mean score decreases and the disease moved from the 1st place to the 8th place. In comparison, FMD ranked high in the prioritization process (2nd), but lowered to the 12th place in the reduced model, which excluded disease pathogen characteristics. For FMD, the strongest driver was the 'pathogens characteristics'. The virus is highly contagious, spreads via airborne and direct contact and affects different livestock species, giving this driver category a strong weight. All experts considered that 'economy and trade activities' was the most important driver (high weight). It was ranked first more often than others. In the reduced model (without the 'economy and trade activities' domain), all diseases with the exception of 7 moved up or down in the ranking by more than 3 places. This is of no surprise, as economic and trade activity has priority in the age of globalization; increased movement of live animals and animal products crossing oceans and international boundaries increase the risk of spread for animal and zoonotic diseases (Domenech, Lubroth, Eddi, Martin, & Roger, 2006 Furthermore, the sensitivity analysis of experts also showed a high correlation among the ranking of models, which confirms that experts were in agreement in regards to the scores. Overall, the importance of validating each generated model is The authors declare no conflict of interest. Ethical statement is not applicable to this study as the data were gathered through questionnaire survey without any animal experimentation. Number of Criteria = 7, hence 70 points to be distributed within this domain for the intra-domain weighing. Mono species farms-One single farmed animal (e.g. only bovines) or multi species farms (farms with more than one species, for example goats and bovines in the same farm/land/premises). Score 0 Score 1 Negligible: the type of farm does not influence in any form (re)emergence of the disease among the livestock population. Score 2 Low: mono or multi species farm has a low effect on the risk of disease to emerge or re-emerge. Score 3 Moderate: the type or types of farmed animals has a moderate effect on the emergence of the disease in Belgium. High: the type of farmed animals has a high influence for the disease to emerge and spread in Belgium. Farm demography/management: such as type of dairy or beef (cattle) production. For pigs-reproduction, fattening, finishing farm or both. Chickens-only laying eggs chickens or solely finishing broilers Score 0 Score 1 Negligible: population demography does not influence in any form the (re)emergence of the disease among the livestock population. Score 2 Low: the demographic population of the farm is a low influencing factor for disease (re)emergence. For example, disease only clinically affects only one age strata (i.e.) newborns, therefore adults are immune to it. Moderate: the demographic of the population has a moderate effect on the (re)emergence of the disease, as it can (re)emerge in more than one type of demography but other conditioning factors have to occur in conjunction. High: the type of demographic of the farm has a high effect on the (re)emergence of the disease as it can (re)emerge in different types of farmed animals and all types of age groups D3 Animal density of farms. Extensive (small holders with a few animals) v/s intensive farming Score 0 Score 1 Negligible: animal farm density is not a risk factor for the disease to emerge in Belgium Score 2 Low: farm density (extensive or intensive) of animals has a low effect on the pathogen's/disease (re)emergence Score 3 Moderate: farm density of animals in the farm (extensive v/s intensive) has a moderate effect on the emergence of pathogen/ disease Score 4 High: farm density of animals has a high effect on the (re)emergence of pathogen/disease. Feeding practices of farms Potential roles of zoo's in the (re)emergence of the pathogen Score 0 Score 1 Negligible: The disease can be present in zoo animals but it is not known to have been transmitted from zoo animals to livestock. Score 2 Low: The disease can enter a zoo (e.g. with introduction of an infected exotic animal) but only accidental transmissions of the disease from zoo animals to livestock have been reported. Hence, zoos have a low effect on the (re)emergence of the disease in Belgium's livestock Score 3 Moderate: The disease can enter a zoo and be present in zoo animals but it needs a vector (biological/mechanical) for its transmission into livestock. Therefore, zoos have a moderate effect on the (re)emergence of the disease in Belgium. High: Disease can be introduced to a zoo via an infected imported animal, zoo animals can carry the disease that can easily jump to livestock animals The rural(farm)-wildlife interface Score 0 Score 1 Negligible: The disease has never (re)emerged from the narrowing of the farm-wild interface Score 2 Low: The disease has a low probability to (re)emerge via the livestock farm-forest interface. The disease has been known to (re)emerge from the wild bush but very rarely Score 3 Moderate: The disease has a moderate probability of (re)emergence via the farm/wildlife interface. Barriers ( natural or artificial) are needed to keep the disease/pathogen (re)emerging in livestock Score 4 High: there is a high probability for the disease to (re)emerge via the farm/forest interface. Barriers (natural or artificial) separating farms from natural forests are ineffective Score 2 Low: there is a low probability of the disease (re)emerging and spreading through increased populations of endemic/migrating wild birds. Disease has spread from the endemic/migrating wild birds but only accidentally or under exceptional circumstances Score 3 Moderate: there is a moderate probability of disease being introduced and spread through increased populations of endemic/migrating wild birds. They are hosts and in close contact with domestic livestock (i.e. poultry farms) may spread the disease Score 4 High: there is a high probability for a disease to (re)emerge through increased populations of wild/migrating birds. These are hosts or reservoirs of the disease Hunting Activities: hunted animals can be brought back to where livestock is present Score 0 Score 1 Negligible: The risk of the disease/pathogen of (re)emerging in livestock due to hunting activities is practically null Score 2 Low: Disease is present in hunted wildlife and birds and only accidental cases have been reported in livestock that have (re)emerged because of hunting. The risk of the disease/pathogen of (re)emerging in livestock due to hunting activities is practically null Score 3 Moderate: Disease is present in hunted wildlife and birds but a certain control is established by the hunter Score 4 High: Disease is present in hunted wildlife and birds and hunting is one of the main modes of transmission of the disease to livestock F6 Transboundary movements of terrestrial wildlife from other countries Score 0 Null: Disease is not carried by terrestrial wildlife Score 1 Negligible: (re)emergence of the disease by terrestrial movements of wildlife has only been suspected but never confirmed. Low: There is a low probability for the disease to (re)emerge and spread through transboundary movements of terrestrial wildlife Score 3 Moderate: There is a moderate probability for the disease to (re)emerge and spread through transboundary movements of terrestrial wildlife Score 4 High: There is a high probability for the disease to (re)emerge and spread through transboundary movements of terrestrial wildlife. These are host and may spread/carry the disease along. Number of Criteria = 6, hence 60 points to be distributed within this domain for the intra-domain weighing. (Continued) G1 In-and out-people movements linked to tourism Score 0 Score 1 Negligible: The movement of tourism is a negligible driver on the emergence or re-emergence of the disease Score 2 Low: Tourism increase has a low driver of the (re)emergence of the disease. Score 3 Moderate: Tourism increase has a moderate driver for the (re)emergence of the disease. Biosecurity measures are enough to stop the entering of the pathogen. High: Tourist movement is a high driver on the (re)emergence of a disease. Tourists are highly likely to bring the disease into Belgium in their belongings and biosecurity measures are insufficient to stop the pathogen G2 Human Immigration Score 0 Score 1 Negligible: The immigration movements are a negligible driver of the disease (re)emergence in Belgium Score 2 Low: The immigration movements are a low driver of the disease (re)emergence in Belgium Score 3 Moderate: The disease is currently present in countries where more immigrants come from and pathogen highly likely to enter through, clothes, shoes and or possession, but the current biosecurity measures in place are able to prevent the emergence of the disease in Belgium Score 4 High: the immigration movement has a high effect as a driver on the emergence or re-emergence of disease in Belgium. Disease is highly likely to emerge using this route as biosecurity measures are not enough to avoid emergence of the disease G3 Transport movements: more specifically commercial flights, commercial transport by ships, cars or military (EXCLUDING TRANSPORT VEHICLES OF LIVE ANIMALS). Score 0 Score 1 Negligible: the role of commercial movements as a driver on the (re)emergence of the disease in Belgium is negligible. Score 2 Low: the role of commercial movements as a driver on the (re)emergence of the disease in Belgium is low. It is easily preventable by implementing biosecurity measures Score 3 Moderate: the role of commercial movements as a driver on the (re)emergence of a disease in Belgium is moderate. Disease can be prevented if biosecurity measures are tightened. High: the role of commercial movements as a driver on the (re)emergence of a disease in Belgium is high. Disease is hard to control via the current biosecurity measures. Transport vehicles of live animals Score 0 Score 1 Negligible: the role of transport vehicles of live animals as a driver for the (re)emergence of the disease in Belgium is negligible Score 2 Low: the role of transport vehicles of live animals as a driver for the (re)emergence of the disease in Belgium is low. Moderate: the role of transport vehicles of live animals as a driver for (re)emergence of the disease in Belgium is moderate. High: the role of transport vehicles of live animals as a driver for (re)emergence of the disease in Belgium is high G5 Bioterrorism potential Score 0 Score 1 Negligible: the role of bioterrorism as a driver for a disease to (re)emerge is negligible: agent is available but difficult to handle or has a low potential of spread or generates few economic consequences Score 2 Low: the role of bioterrorism as a driver for a disease to (re)emerge is low: agent is available and easy to handle by professionals and labs but has a low spread Score 3 Moderate: the role of bioterrorism as a driver for a disease to (re)emerge is moderate: agent available and easy to handle by professionals and labs and rapidly spreads Score 4 High: the role of bioterrorism as a driver for a disease to (re)emerge is high: Agent is available and easy to handle by individuals and rapidly spreads G6 Inadvertent release of an exotic infectious agent from a containment facility, for example Laboratory Score 0 Score 1 Negligible: the pathogen is not currently present in any laboratory Score 2 Low: the pathogen is present in a containment facility but its release is very unlikely as it is very easily contained Score Score 0 Score 1 Negligible: modification of the disease status due to a reduced national budget has a negligible effect on the (re) emergence of the disease in Belgium Score 2 Low: modification of the disease status due to a reduced national budget has a low effect on the (re) emergence of the disease in Belgium Score 3 Moderate: modification of the disease status due to a reduced national budget has a moderate effect on the (re) emergence of the disease in Belgium Score 4 High: modification of the disease status due to a reduced national budget has a high effect on the (re) emergence of the disease in Belgium Decrease of resources allocated to the implementation of biosecurity measures at border controls (e.g. harbours or airports). Score 0 Score 1 Negligible: decreasing the resources allocated to the implementation of biosecurity measures has a negligible effect on the (re)emergence of the disease in Belgium. Disease has never been detected in the past in a harbour or airport Score 2 Low: decreasing the resources allocated to the implementation of biosecurity measures has a low effect on the (re)emergence of the disease in Belgium. The disease has been suspected to have entered other countries because of deficient biosecurity at border controls. Score 3 Medium: decreasing the resources allocated to the implementation of biosecurity measures has a moderate effect on the (re) emergence of the disease in Belgium. The disease has been introduced in other countries because of deficient biosecurity at border controls Score 4 High: decreasing the resources allocated to the implementation of biosecurity measures highly increases the risk of (re)emergence of the disease in Belgium. In the past, the disease has been introduced in other countries AND in Belgium because of deficient biosecurity at border controls Most likely influence of (il)legal movements of live animals (livestock, pets, horses, etc.) from neighbouring/European Union member states (MS) for the disease to (re)emerge in Belgium. (Continued) H5 Influence of increased (il)legal imports of animal subproducts such as skin, meat and edible products from EU member states for the disease/pathogen to (re)emerge in Belgium Score 0 Score 1 Negligible: increased (il)legal imports of animal subproducts such as skin, meat and edible products from EU member states have a negligible influence on the pathogen/disease (re)emergence in Belgium. Score 2 Low: increased (il)legal imports of animal subproducts such as skin, meat and edible products from EU member states have a low influence on the pathogen/disease (re)emergence in Belgium. Moderate: increased (il)legal imports of animal subproducts such as skin, meat and edible products from EU member states have a moderate influence on the pathogen/disease (re)emergence in Belgium. High: increased (il)legal imports of animal subproducts such as skin, meat and edible products from EU member states have a high influence on the pathogen/disease (re)emergence in Belgium. Most likely influence of increased (il)legal imports of NON-animal products such as tires, wood, furniture from EU member states for the disease/pathogen to (re)emerge in Belgium. Score 0 Score 1 Negligible: increased (il)legal imports of NON-animal products such as tires, wood, furniture from EU member states have a negligible influence on the pathogen/disease (re)emergence in Belgium. Score 2 Low: increased (il)legal imports of NON-animal products such as tires, wood, furniture from EU member states have a low influence on the pathogen/disease (re)emergence in Belgium. Moderate: increased (il)legal imports of NON-animal products such as tires, wood, furniture from EU member states have a moderate influence on the pathogen/disease (re)emergence in Belgium. High: increased (il)legal imports of NON-animal products such as tires, wood, furniture from EU member states have a high influence on the pathogen/disease (re)emergence in Belgium. Most likely influence of (il)legal movements of live animals (livestock, pets, horses, etc.) from Third countries for the disease to (re)emerge in Belgium. Most likely influence of increased imports of animal subproducts such as skin, meat and edible products from Third countries, for the disease to (re)emerge in Belgium. Score 0 Score 1 Negligible: Increased imports of animal subproducts such as skin, meat and edible products from Third countries have a negligible influence on the pathogen/disease (re)emergence in Belgium. Score 2 Low: Increased imports of animal subproducts such as skin, meat and edible products from Third countries have a low influence on the pathogen/disease (re)emergence in Belgium. Moderate: Increased imports of animal subproducts such as skin, meat and edible products from Third countries have a moderate influence on the pathogen/disease (re)emergence in Belgium. High: Increased imports of animal subproducts such as skin, meat and edible products from Third countries have a high influence on the pathogen/disease (re)emergence in Belgium. A P P E N D I X A (Continued) H9 Most likely influence of increased (il)legal imports of NON-animal products such as tires, wood, furniture from Third countries, for the disease to (re)emerge in Belgium. Score 0 Score 1 Negligible: increased (il)legal imports of NON-animal products such as tires, wood, furniture from Third countries have a negligible influence on the pathogen/disease (re)emergence in Belgium. Score 2 Low: increased (il)legal imports of NON-animal products such as tires, wood, furniture from Third countries have a low influence on the pathogen/disease (re)emergence in Belgium. Moderate: increased (il)legal imports of NON-animal products such as tires, wood, furniture from Third countries have a moderate influence on the pathogen/disease (re)emergence in Belgium. High: increased (il)legal imports of NON-animal products such as tires, wood, furniture from Third countries have a high influence on the pathogen/disease (re)emergence in Belgium. Number of Criteria = 9, hence 90 points to be distributed within this domain for the intra-domain weighing. A P P E N D I X B List of experts enrolled (N = 14) in the phase I (questionnaire assessment) with their gender, affiliation, country and field of expertise The Schmallenberg virus epidemic in Europe-2011-2013. Preventive Veterinary Medicine Molecular characterization of Peste des petits ruminants viruses in the Marmara Region of Turkey Climate change and infectious diseases: From evidence to a predictive framework Schmallenberg virus continues to spread across Europe. The Veterinary Record Antibodies to Schmallenberg virus in domestic livestock in Turkey Surveillance, detection and response: Managing emerging diseases at national and international levels Building a picture: Prioritisation of exotic diseases for the pig industry in Australia using multi-criteria decision analysis Evidence-based semiquantitative methodology for prioritization of foodborne zoonoses African horse sickness virus: History, transmission, and current status Culicoides and the emergence of bluetongue virus in northern Europe Avian influenza. Iowa State University Evidence of Schmallenberg virus circulation in ruminants in Greece Identification of wild boar-habitat epidemiologic cycle in African Swine fever epizootic Multi-criteria decision analysis tools for prioritising emerging or re-emerging infectious diseases associated with climate change in Canada Emerging infectious diseases of wildlife-threats to biodiversity and human health Regional and international approaches on prevention and control of animal transboundary and emerging diseases The highly pathogenic avian influenza A (H7N7) virus epidemic in The Netherlands in 2003-lessons learned from the first five outbreaks amending Annexes I and II to Council Directive 82/894/EEC on the notification of animal diseases within the Community Scientific Opinion on African swine fever Scientific opinion on porcine epidemic diarrhoea and emerging porcine deltacoronavirus Situation zoosanitaire et maladies à déclaration obligatoire en Belgique Notification Obligatoire Federal Ministry of Food and Agriculture of Germany Animal production health Phylogeographic analysis of African Swine Fever Virus Prioritizing emerging zoonoses in the Netherlands Multidisciplinary and evidence-based method for prioritizing diseases of food-producing animals and zoonoses Emerging and re-emerging zoonotic diseases: Challenges and opportunities. In 72nd General session World Organisation for Animal Health Outbreak of foot-and-mouth disease virus serotype O in the UK caused by a pandemic strain Le service public de l'accès au droit Le service public de l'accès au droit Schmallenberg virus infection in small ruminants -First review of the situation and prospects in Northern Europe Summer 2018: African swine fever virus hits north-western Europe Nederlandse Voedsel-en Warenautoriteit Emergence of a highly pathogenic avian influenza virus from a low-pathogenic progenitor The 2010 foot-and-mouth disease epidemic in Japan Unexpected discovery of African swine fever in Belgium West Nile virus in Europe: Emergence, epidemiology, diagnosis, treatment, and prevention List of notifiable diseases Great Britain Complete genome sequence of a porcine epidemic diarrhea virus from a novel outbreak in Belgium Bluetongue in Europe: Past, present and future animal-health-in-the-world/ oie-listed-disea ses-2018/ World Organisation for Animal Health Detection and partial sequencing of Schmallenberg virus in cattle and sheep in Turkey. Vector-Borne and Zoonotic Diseases Contagious caprine pleuropneumonia 3,617.45 (1,099.65) 11 3,247.25 21 2 Epizootic haemorrhagic disease 3 African swine fever 3