imprimir | regrese al sitio web


Archivo: 1999 Session - Appendix 1

1999 Session of the Research Group of the Standing Technical Committee of EuFMD

Risk and economic consequences of exotic animal disease outbreaks: Computer simulation to help set priorities in policy making

Dr. Suzan Horst Wageningen University, Farm Management


Outbreaks of OIE-list A diseases, such as Foot and Mouth Disease (FMD) and Classical swine Fever (CSF) will have severe economic consequences for countries with a high density of animals and farms and an export-oriented production. For such countries, effective disease prevention and eradication programs for exotic animal disease are of great importance. Historical and experimental data to base these programs on is scarce. Extensive research has been done, but mainly focused on the biological aspects of virus and virus spread, not so much on virus introduction. Also an integrated approach which combines the various aspects of outbreaks and quantifies the risks and economic consequences is lacking. Within the Animal Health Economics section of the Farm Management group of the Wageningen University, research projects are carried out aimed at providing more insights into this issue. The projects often result in information and insights that might be useful as bricks to build upon sound decision making in the area of disease prevention and eradication. Because for diseases such as FMD, large scale real life experiments are too risky, computer simulations models are used as an alternative. The presentation will give an overview of the results of the already completed research project named ÍRisk and economic consequences of contagious animal disease introductionÔ (Horst, 1998). Thereafter, attention will be given to recent work concerning spread of contagious diseases within a country. Finally, a short overview of ongoing projects will be provided. The underlying paper gives a summary of the project ÍRisk and economic consequences of contagious animal disease introductionÔ.



The project ÍRisk and economic consequences of contagious animal disease introductionÔ was the result of concerted action by the government and private industry aimed at gaining more insight into the risk and economic consequences of virus introduction into the Netherlands. The project resulted in a general outline for an integrated approach, described by Horst et al. (1996). Based on this approach, a simulation model was developed to evaluate the process of virus introduction from European countries into the Netherlands and to identify the key risk factors. The model, called VIRiS (Virus Introduction Risk Simulation) focuses on CSF and FMD and is based on historical and experimental data, where available and applicable. Expert knowledge was required and was used as an additional source of information. To rate the information obtained by the VIRiS-model at its true, the results obtained were combined with information from models describing the spread of virus within a country (Jalvingh et al., 1996; Saatkamp, 1996) and thereafter translated into their economic consequences. In this way alternative strategies to prevent introduction of virus in the Netherlands could be evaluated on their merits concerning reduction of losses due to outbreaks.





The VIRiS model

The VIRiS model is based on Monte Carlo sampling. This type of modelling uses randomly drawn numbers from specified distributions and therefore acknowledges the fact that there is variation in the system to be modelled. This variation is due to uncertainty of the exact value of the input parameters and to the nature of the system itself, i.e., also in reality outbreaks may vary in duration, number of outbreaks may differ between years. When such a Monte Carlo simulation model is run repeatedly, the distribution of output values will realistically reflect the possible behaviour of the system under study.

A complete VIRiS run has a time horizon of 5 years. The model uses time steps of one day. Each ÍdayÔ, the model starts with sampling for new outbreaks in the countries in Europe. If in one of these countries (grouped into five clustered for pragmatic reasons) an outbreak is sampled, the model ÍfollowsÔ the disease up to a possible introduction of virus into the Netherlands. Distributions based on quantitative information on epidemiology, animal density and trade flows determine the location of the possible outbreak in the Netherlands. The number, location and cause (country and method of transmission) of occurring Dutch outbreaks are stored. Intermediate year results are calculated. Days and years are linked to each other in order to follow the evolution of outbreaks.
Input for VIRiS: elicitation of expert knowledge
Three major aspects determine the risk of virus introduction:


1) Frequency of outbreaks in European countries;
A higher outbreak frequency in countries with which the Netherlands has any kind of relationship (e.g. trade) results in a higher introduction risk for the Netherlands


2) Duration of the High Risk period
The High Risk period (HRP) defines the period in which virus is already present in a country but not yet detected or under control. During this period virus might be transferred easily to another country.



3) Risk factors
Risk factors are the vehicles used for virus transfer. Examples of risk factors are: import of livestock, import of animal products, and returning empty trucks (used for livestock export).
Information on these three aspects was obtained from scientific literature, statistics of the National Product Boards, and databases such as the Animal Disease Notification System of the EU. However, changing circumstances such as increased free-trade within the EU, increasing importance of Eastern European countries as trade partner, and ongoing developments in animal health programs and animal health legislation make it difficult to base assumptions and estimates on historical data only. Therefore additional information was sought by experts, people involved in the development and execution of disease eradication and prevention programs. Their expertise combines experience and understanding of current and future circumstances and was therefore seen as a useful and necessary addition.
To elicit and quantify expert information, workshops were organised during which the participants were asked to complete a computerised questionnaire. The questionnaire was self-explanatory in order to minimise the interaction of the participants with each other of with the workshop facilitators. Several elicitation techniques were used, among others the so-called ÍConjoint Analysis techniqueÔ (CA). CA is widely used in market research to elicit the subjective views and preferences of consumers (Cattin & Wittink, 1982). The technique is based on the assumption that a product or event can be evaluated as a package of characteristics or attributes and that the overall judgement of the package is based on the individual importance of each of these attributes. CA starts with the overall judgements of respondents on a set of alternatives (i.e., combinations of attributes) and breaks these overall judgements down into the contribution of each attribute, by using ordinary least-squares (OLS) regression analysis. The technique was used in this study to elicit the relative importance of the risk factors. A detailed description of the workshop experiment is given by Horst et al. (1997). Because of the positive evaluation, the same technique was used in a Swiss experiment, aimed at the elicitation of knowledge on introduction and spread of CSF in Switzerland, described by St®rk (1997).



Spread of virus within the country


Models of Jalvingh et al. (1996) and Saatkamp (1996) describe the consequences of primary outbreaks of restively FMD and CSF in the Netherlands. The blue print of the FMD model was developed in New Zealand. Jalvingh et al. (1996) adapted the model to Dutch circumstances. As VIRiS, also InterSpread is a Monte Carlo simulation model. Saatkamp (1996) developed a state-transition model (with some Monte Carlo elements) describing the spread of CSF within Belgium. The model was adapted to Dutch circumstances. Both Íspread modelsÔ provide insight into the number of infected and affected farms (affected by movement restrictions), and the duration of the outbreak, under various eradication strategies. In this study, the minimal EU standard strategy was applied.



Economic consequences of outbreaks

In close corporation with representatives from the livestock production sector, losses due to outbreaks of CSF and FMD were calculated, for all parts of the production chain. Losses were due to slaughtering of infected herds, production stand-still (farms, slaughterhouses), movement restrictions, buying-up measures (government buys and destroys animals from farms within the restriction zones, to avoid problems with welfare, virus transfer and oversupply on the internal market due to trade bans), and organisational costs.




VIRiS results show that most primary CSF outbreaks are expected to occur in the southern region of the Netherlands (41%) followed by the eastern region (34%). Least outbreaks are expected in the northern and western region, 12 and 13% respectively. The most likely area in the Netherlands for a primary FMD outbreak is the eastern part (38%). In the southern, western and northern part the expected percentage of primary outbreaks in 26, 22, and 14% respectively. These probabilities have a high and positive correlation with the animal and farm density of the respective areas, which is due to the great influence of the risk factor Íimport of livestockÔ. More than 50% of the expected primary FMD outbreaks were caused by this factor. Also for CSF the import of livestock is the most important risk factor. Other important risk factors are Íreturning trucksÔ (both diseases) and Íswill feedingÔ (CSF). Most outbreaks of both disease are originating from southern European countries or from countries neighbouring the Netherlands.

Table 1 shows the results of the base and some alternative scenarios of the VIRiS model. The base scenario describes the current situation in the Netherlands, based on the assumption that, most likely, 2.5 CSF and 1 FMD outbreak will occur in the next five year. The calculations are based on the average duration and magnitude of outbreaks (results of the Íspread modelsÔ). Outbreaks were stamped out in accordance with minimal EU-standards. The results show that on average, expected annual losses due to CSF outbreaks are 77 million Dutch guilders (about US$ 39 million). Expected annual losses due to FMD outbreaks are lower because the frequency of such outbreaks is lower and also losses per outbreak are lower (duration of outbreaks is shorter).

The alternative scenarios show the importance of early detection of outbreaks. Doubling HRP1 (interval infection - detection) results in a considerable increase in the expected annual losses. The complete elimination of the risk associated with returning trucks results in a reduction of expected annual losses of more than 10 million Dutch guilders. The reduction in losses, obtained by the alternative scenarios, might be seen as an indication for the financial space available for measures aimed at the realisation of the respective scenarios. In this way, VIRiS can add to the evaluation of animal disease prevention programs and serve as a useful tool for policy makers.

See Table 1. Number of outbreaks per year and losses per year (in million Dutch guilders)




Simulation modelling is a powerful technique to gain insight into complex problem situations. The described VIRiS model is flexible and can easily be adapted to new situations and new information. The current model is partly based on expert information. As long as historical data and/or experimental research are not available to provide better predictive results, the use of such expert information will be necessary in models describing risks and consequences of contagious animal disease outbreaks. The Monte Carlo technique enables an optimal use of this information. The study shows that the integration of models describing introduction, spread and economic consequences of disease outbreaks provides useful information for the evaluation of animal disease prevention programs.



  • Cattin P, Wittink D.R., 1982. Commercial use of conjoint analysis: a survey. Journal of Marketing, 46, 44-53.
  • Horst, H.S. 1998. Risk and economic consequences of contagious animal disease introduction. Manholt Studies 11, Backhuys Publishers, Leiden, the Netherlands, 145 pp.
  • Horst H.S., Dijkhuizen A.A., Huirne R.B.M., 1996. Outline for an integrated modelling approach concerning risks and economic consequences of contagious animal disease. Netherlands Journal of Agricultural Science 44, 89-102.
  • Horst H.S., Dijkhuizen A.A., Huirne R.B.M., De Leeuw P.W., 1997. Introduction of contagious animal diseases into the Netherlands: elicitation of expert opinion. livestock Production Science. In press.
  • Jalvingh A.W., Stern M.W., Dijkhuizen A.A., Morris R.S., 1996. Stochastic and spatial simulation of contagious animal disease outbreaks. Proceedings ICCTA Í96. Agro-informaticareeks 10, 305-309.
  • Saatkamp H.W., 1996. Simulation studies on the potential role of national identification and recording systems in the control of classical swine fever. Mansholt Studies 2, Backhuys Publishers, Leiden, 120 pp.
  • St®rk K.D.C., Horst H.S., Morris R.S., Teuffert J., 1997. Elicitation of expert knowledge on risk factors for classical swine fever transmission in Switzerland. Proceedings of the VIIIth Symposium of the International Society for Veterinary Epidemiology and Economics, Paris, in press.