Risk factors for Campylobacter colonization in Danish broiler flocks, 2010 to 2011.

The objectives of the two studies presented were to estimate the prevalence of Campylobacter-positive farms and flocks and to acquire updated knowledge about risk factors for the introduction of Campylobacter in Danish broiler flocks. In the first study, from September 2010 to September 2011, there were 25 Danish broiler farms visited, and a questionnaire was filled in by a veterinarian/consultant. In the second study, a similar questionnaire was distributed electronically to all Danish broiler farmers (n=164) that were on record with an email address in the Quality Assurance System in the Danish Broiler Production (KIK) database. House- and flock-specific data collected in the surveys were supplemented with information obtained from the KIK database. Data obtained from the two studies were analyzed separately by logistic regression analysis. In both models, the dependent variable was "Campylobacter flock status (positive/negative)," which was based on real-time PCR testing of fecal material from the floor of each broiler house that had been collected preslaughter using a pair of tube gauze "socks." This material was pooled into one sample. Of the 25 farms visited, 17 had delivered Campylobacter-positive flocks during the study period, and eight farms had no Campylobacter-positive flocks. Moreover, the flock prevalence of Campylobacter was 17% (n=418). Data obtained from the electronically distributed survey revealed that 63% (n=71) of the farms were Campylobacter-positive. Further, the flock prevalence of Campylobacter was 14% (n=1,286). The multivariable models from the two sets of data identified the following statistically significant risk factors: summer vs. winter; if the previous flock in the house was positive for Campylobacter vs. if the previous flock in the house was negative; and litter delivered into the house close to the time of arrival of new chickens vs. storing litter on the farm. Furthermore, the data showed that a vertically based ventilation system (where most of the air is taken into the house through chimneys) constituted a higher risk than a horizontally based system (where the air is taken in through openings in the side walls).

Multidetector CT features of mesenteric vein thrombosis.

Mesenteric vein thrombosis (MVT) accounts for 5%-15% of all mesenteric ischemic events and is classified as either primary or secondary. Primary MVT is idiopathic, whereas secondary MVT can result from a variety of underlying diseases and risk factors, including primary hypercoagulable states or prothrombotic disorders, myeloproliferative neoplasms, cancer (most frequently of the pancreas or liver), diverse inflammatory conditions, recent surgery, portal hypertension, and miscellaneous causes such as oral contraceptives or pregnancy. Clinical symptoms of MVT are rather nonspecific and are mainly characterized by abdominal pain. The mortality rate for MVT remains high, since even now the diagnosis is often delayed. Multidetector computed tomography (CT) is the modality of choice in this context. Although venous bowel ischemia occurs only infrequently with MVT, radiologists should be familiar with its multidetector CT features. Familiarity with the possible causes of MVT, the underlying pathogenic mechanisms associated with MVT, and the correlation between multidetector CT features and these pathogenic mechanisms is necessary to optimize medical management and improve patient care.

Modelling food safety and economic consequences of surveillance and control strategies for Salmonella in pigs and pork.

Targets for maximum acceptable levels of Salmonella in pigs and pork are to be decided. A stochastic simulation model accounting for herd and abattoir information was used to evaluate food safety and economic consequences of different surveillance and control strategies, based among others on Danish surveillance data. An epidemiological module simulated the Salmonella carcass prevalence for different scenarios. Cost-effectiveness analysis was used to compare the costs of the different scenarios with their expected effectiveness. Herd interventions were not found sufficient to attain Salmonella carcass prevalence <1%. The cost-effectiveness of abattoir interventions changed with abattoir size. The most cost-effective strategy included the use of steam vacuum and steam ultrasound. Given uncertainty of the effect of steam vacuum and steam ultrasound, model results should be updated as more information becomes available. This framework contributes to informed decision-making for a more cost-effective surveillance and control of Salmonella in pigs and pork.

Quantitative assessment of the likelihood of the introduction of classical swine fever virus into the Danish swine population.

Classical swine fever virus (CSFV) is a major infectious-disease agent of livestock and causes production losses through increased morbidity and mortality, particularly of young pigs. We identified the pathways for introduction of CSFV into Denmark and assessed the annual probability of introduction (based on a US Department of Agriculture model). We developed pathways based on material from scientific articles, reports from veterinary agencies and custom officers, and consultations with experts in the field. Returning livestock trucks and legal meat imports were the most important pathways for CSFV introduction to Denmark from other EU states with predicted overall likelihood of one or more introductions of CSFV within a median of 130 years (46-280) provided mitigating steps, such as cleaning trucks, were maintained to a very high standard. The likelihood would increase dramatically if these activities were abandoned: one or more introductions within a median of 5.2 years (2-14). The predicted risks from live-animal imports and semen were extremely low given the very few imports of these products. The most important countries for Denmark's CSFV risk are Germany and the Netherlands, though this risk is again predicted to be dramatically reduced as long as mitigating activities are maintained. We predicted the risk from illegal movements of pork into Denmark to be low because little pork enters through this route and only a small fraction of this pork would be fed to pigs.

Simulating the spread of classical swine fever virus between a hypothetical wild-boar population and domestic pig herds in Denmark.

Denmark has no free-range wild-boar population. However, Danish wildlife organizations have suggested that wild boar should be reintroduced into the wild to broaden national biodiversity. Danish pig farmers fear that this would lead to a higher risk of introduction of classical swine fever virus (CSFV), which could have enormous consequences in terms of loss of pork exports. We conducted a risk assessment to address the additional risk of introducing and spreading CSFV due to the reintroduction of wild boar. In this paper, we present the part of the risk assessment that deals with the spread of CSFV between the hypothetical wild-boar population and the domestic population. Furthermore, the economic impact is assessed taking the perspective of the Danish national budget and the Danish pig industry. We used InterSpreadPlus to model the differential classical swine fever (CSF) risk due to wild boar. Nine scenarios were run to elucidate the effect of: (a) presence of wild boar (yes/no), (b) locations for the index case (domestic pig herd/wild-boar group), (c) type of control strategy for wild boar (hunting/vaccination) and (d) presence of free-range domestic pigs. The presence of free-range wild boar was simulated in two large forests using data from wildlife studies and Danish habitat data. For each scenario, we estimated (1) the control costs borne by the veterinary authorities, (2) the control-related costs to farmers and (3) the loss of exports associated with an epidemic. Our simulations predict that CSFV will be transmitted from the domestic pig population to wild boar if the infected domestic pig herd is located close to an area with wild boar (<5 km). If an outbreak begins in the wild-boar population, the epidemic will last longer and will occasionally lead to several epidemics because of periodic transfer of virus from groups of infected wild boar to domestic pig herds. The size and duration of the epidemic will be reduced if there are no free-range domestic pig herds in the area with CSF-infected wild boar. The economic calculations showed that the total national costs for Denmark (i.e. the direct costs to the national budget and the costs to the pig industry) related to an outbreak of CSF in Denmark will be highly driven by the reactions of the export markets and in particular of the non-EU markets. Unfortunately, there is a substantial amount of uncertainty surrounding this issue. If hunting is used as a control measure, the average expenses related to a CSF outbreak will be 40% higher if wild boar are present compared with not present. However, a vaccination strategy for wild boar will double the total costs compared with a hunting strategy.

Risk of rabies introduction by non-commercial movement of pets.

In response to a Commission request, EFSA has carried out a quantitative assessment of the risk of rabies introduction into the UK, Ireland, Sweden, and Malta due to the movement of pets incubating rabies at the time of movement. The risk that a pet is incubating rabies at the time of first vaccination is equal to the prevalence of rabies-incubating pets in the population of origin. Following induction of protective immunity by vaccination, animals already incubating rabies will still develop clinical disease as a function of time after vaccination (termed type A risk). A waiting period will reduce this risk. Afew animals may not be protected after single-shot primary vaccination. Such animals may become infected during the waiting period after vaccination. The risk of becoming infected after the first vaccination (termed type B risk) depends on the prevalence and efficiency of vaccination. Serological testing can be used to identify non-immune pets (depending on test specificity) and will therefore reduce this risk accordingly. The type A and B risks were modelled as a function of the waiting period after vaccination and fitted to a non-linear model incorporating vaccination efficiency and test specificity. The model can be used to quantify the risk of moving pets from rabies infected areas and also to investigate the effect of different control measures. In quantitative terms, the type A risk constitutes by far the major risk. Therefore, a waiting period (defined as the time spent between vaccination and pet movement to the destined country) is the major effective measure to mitigate the risk of rabies introduction due to an animal being infected before primo-vaccination. Serological testing will only add significantly to risk reduction when waiting periods exceed 100 days. Within the EU, the rabies prevalence in most countries is so low that the risk can be considered negligible. However, for some countries the risk is non-negligible.

Review of transmission routes of 24 infectious diseases preventable by biosecurity measures and comparison of the implementation of these measures in pig herds in six European countries.

This study aimed to review the transmission routes of important infectious pig diseases and to translate these into biosecurity measures preventing or reducing the transmission between and within pig herds. Furthermore, it aimed to identify the level of implementation of these measures in different European countries and discuss the observed variations to identify potentials for improvement. First, a literature review was performed to show which direct and indirect transmission routes of 24 infectious pig diseases can be prevented through different biosecurity measures. Second, a quantitative analysis was performed using the Biocheck.UGent™, a risk-based scoring system to evaluate biosecurity in pig herds, to obtain an insight into the implementation of these biosecurity measures. The database contained farm-specific biosecurity data from 574 pig farms in Belgium, Denmark, France, Germany, the Netherlands and Sweden, entered between January 2014 and January 2016. Third, a qualitative analysis based on a review of literature and other relevant information resources was performed for every subcategory of internal and external biosecurity in the Biocheck.UGent™ questionnaire. The quantitative analysis indicated that at the level of internal, external and overall biosecurity, Denmark had a significantly distinct profile with higher external biosecurity scores and less variation than the rest of the countries. This is likely due to a widely used specific pathogen-free (SPF) system with extensive focus on biosecurity since 1971 in Denmark. However, the observed pattern may also be attributed to differences in data collection methods. The qualitative analysis identified differences in applied policies, legislation, disease status, pig farm density, farming culture and habits between countries that can be used for shaping country-specific biosecurity advice to attain improved prevention and control of important pig diseases in European pig farms.

Estimation of Salmonella prevalence on individual-level based upon pooled swab samples from swine carcasses.

Pooling of samples might be an effective means to increase cost-effectiveness in routine surveillance. The present study assessed the effect on the sensitivity of detection of Salmonella when pooling swab samples from swine carcasses compared to individual analyses. A total of 18,984 samples from nine Danish swine abattoirs were collected during 1 year, covering 2017 slaughter days. At each abattoir, swab samples were taken on a daily basis from 10 carcasses randomly selected. From each carcass, an area of 3 cm x 100 cm was swabbed. Five of these samples were analysed individually and the other five were analysed as one pooled sample. Standard culture methods were used. A logistic regression model was built, where the response was whether a sample was Salmonella positive or not. The explanatory factors were abattoir, type of sampling (individual or pooled sample), and season of year 2000 (four quarters). The odds ratio (OR) of the effect of type of sampling in the logistic model accounting for abattoir and season was interpreted as the conversion factor between pooled and individual sample prevalence. The results of the individually analysed samples showed a low prevalence of Salmonella (1.4%). When Salmonella was isolated, mostly only one positive sample was found among the five individually analysed samples per slaughter day. On a few days >1 positive samples' were found (9 out of 2017 days approximately 0.4%). The pooled sample prevalence was 4.1%. Because the individual prevalence was low, the pooled sample prevalence would have been around five times higher than the individual-level prevalence-if there had been no loss of sensitivity. However, we found that due to loss of sensitivity the pooled prevalence was only three times higher (OR = 2.7; CI 2.0-3.7). Therefore, a conversion factor of 3 instead of 5 should be applied to calculate the individual prevalence from a pooled prevalence. This approach has been used in the national surveillance of Danish pork since 2001. The estimated conversion factor and accept of pooling samples do not necessarily apply to a population with a higher prevalence or to other types of samples (e.g. faeces or lymph nodes) or diagnostic procedures.

Where should the effort be put to reduce the Salmonella prevalence in the slaughtered swine carcass effectively?

Infection with Salmonella Typhimurium is seldom associated with clinical disease in pigs. However, control is important as the public is concerned about the human health impact. The producers, the abattoirs and the authorities are interested in implementing procedures to mitigate this risk. To evaluate the effect of different procedures, a stochastic risk model was developed to simulate the prevalence of Salmonella infection during the production process from the live pig on the farm, to the final carcass. This paper describes the model and findings of simulating different control scenarios. The variables with maximum effect on the Salmonella prevalence on the final carcass were (1) number of herds with a high prevalence of Salmonella, (2) singeing efficiency, (3) contamination and cross-contamination at degutting and (4) cross-contamination during handling. However, improvement of any single factor in isolation had a limited impact upon the level of contamination. The largest reduction was observed when several factors were improved concurrently.

Survival of Escherichia coli and Salmonella Typhimurium in slurry applied to clay soil on a Danish swine farm.

A pilot study was carried out on a Danish swine farm infected with multi-resistant Salmonella Typhimurium DT104 (MRDT104). We aimed to (1) investigate to which degree the decline of Escherichia coli and Salmonella in swine slurry applied to farmland depended on the application method; (2) estimate the survival times of E. coli and Salmonella in the soil surface following deposition of naturally contaminated pig slurry; and (3) simulate survival of Salmonella in different infection levels using E. coli data as input estimates. Slurry was deposited by four different methods: (1) hose applicator on black soil followed by ploughing and harrowing; (2) hose applicator on black soil followed only by harrowing; (3) hose applicator on a field with winter-wheat seedlings without further soil treatment; (4) slurry injector on a field with winter-wheat seedlings without further soil treatment. E. coli and Salmonella could not be detected at all in soil following treatment 1. Following the other treatments, E. coli was not detected in soil samples after day 21 and Salmonella was no longer detected after day 7. Simulation results showed that clinical (4 log CFU g(-1)) and sub-clinical Salmonella levels (2500 CFU g(-1)) would fall below the detection limit within 10 or 5 days, respectively. Analysis of samples from 62 Danish MRDT104-infected swineherds showed that nearly 75% of these herds had low levels of MRDT104 (< 10 CFU g(-1)) in their slurry. Our results show that ploughing and harrowing of soil amended with contaminated pig slurry was an effective means to reduce environmental exposure to E. coli and Salmonella on this clay-soil farm.

The correlation between Salmonella serology and isolation of Salmonella in Danish pigs at slaughter.

In Denmark, a serological Salmonella surveillance programme in finishing pig herds has been in place since 1995. The programme was founded on data from experimental studies, which demonstrated a strong association between Salmonella serology and the prevalence of these bacteria. The current study was carried out in three Danish abattoirs to evaluate the correlation under field conditions. A total of 160 Danish finishing pig herds were included. Seven out of these were examined twice, yielding a total of 167 observations. The herds were selected according to their herd serology based on data from the national surveillance. From each herd, samples were taken from 10 finishers at slaughter. The prevalence of Salmonella bacteria was measured at four sites: (1) caecal-content; (2) carcass surface; (3) pharynx; and (4) caecal lymph nodes. A logistic regression model was constructed for each sampling site. Abattoir, sanitary slaughter and herd seroprevalence were used as explanatory variables. The results demonstrated that there was a strong association between herd serology and the prevalence of Salmonella bacteria measured at three of the sampling sites: caecal-content, pharynx, and carcass surface. For these sites, the odds for being culture-positive for Salmonella varied from 1.3 to 1.5 for each increase of 10% in herd serology (P < 0.0001). For caecal lymph nodes, however, no linear association was found.

Biosecurity in 116 Danish fattening swineherds: descriptive results and factor analysis.

Biosecurity measures are important for the herd's protection against diseases and also to provide nationwide protection against the introduction of exotic diseases. In this paper, we describe the farmers' choices and routines regarding biosecurity in Danish fattening herds. Overall, 116 Danish swine fattening herds in three areas of different pig density were interviewed. Of these herds, 78% purchased weaners from one sow herd only, whereas 10% purchased from >5 sow herds during 1 year. Large herds (>500 finishers) purchased weaners more often than smaller herds (< or =500 finishers). Only two swine producers purchased weaners from a market. Almost all herds (95%) received weaners from one sow herd at one time, and only one herd received from >5 sow herds in one batch. Twenty-one percent did not use an effective barrier between the loading area and the stables when delivering pigs for slaughter. Entry rooms (in which clothing and boots are changed) were common at the farm, and the numbers of visitors were generally low (<10 visitors a year). Factor analysis was performed to find underlying structures of the biosecurity measures in the herd; four factors were chosen (eigenvalue > 3.10). A site scoring high on factor 1 was a large SPF herd, which received weaners from a single source, had biosecurity requirements for the transport vehicles, and had a high level of biosecurity for visitors. A site scoring high on factor 2 was a multi-site farm, which had personnel working on more than one of the sites, only received weaners from one sow herd, had delivering herds placed close to the participating site, and transported animals themselves. A site scoring high on factor 3 was a site which hired commercial transport for slaughter, was situated far from the abattoir and had a high level of biosecurity when loading pigs. A production site scoring high on factor 4 was a large site, which used all-in/all-out management, washed and disinfected between each group, and purchased many weaners.

A cross-sectional study of paratuberculosis in 1155 Danish dairy cows.

In a cross-sectional study on milk samples from 1155 cows from 22 Danish dairy herds, selected risk factors for paratuberculosis were identified. The diagnostic procedure used was an indirect enzyme-linked immuno-sorbent assay (ELISA) to detect antibodies against Mycobacterium avium subsp. paratuberculosis. A sample was considered test-positive if it had a corrected optical density >/=0. 025 (test sensitivity 71.4% and test specificity 89.7%). Of the 1155 samples, 8.8% (102/1155) were test-positive, and 19 out of the 22 dairy herds had >/=1 test-positive cows. The significant risk factors in a multiple logistic regression analysis were: Jersey versus large breeds, high parity versus low parity, the first month after calving versus other months of lactation, and a large herd size compared to a small herd size. The highest probability (37-38%) of a positive test was observed among older cows (parity >4) and tested within the first month after calving (irrespective of breed). The lowest probability (2%) of a positive test-result was observed among first parity, large-breed cows tested before calving or later than one month after.

Risk factors for metritis in Danish dairy cows.

A retrospective longitudinal study of metritis was conducted in Denmark on data collected during 1993-1994. Data on herd size, breed, parity, and treatment of disease were obtained from the Danish Cattle Database. Management and production-facility data were collected using a questionnaire, conducted as a telephone interview in 1994. The study included 2144 herds from three regions in Denmark (102,060 cows). Herd-level variables included were: herd size, housing, flooring, grazing, calving measures, and calving supervision. Cow-level variables were: parity, breed, calving season and whether the cow had been treated by a veterinarian for dystocia or the diseases: retained placenta, reproductive disease, ketosis, milk fever, or dry cow mastitis. Marginal multivariable logistic-regression analyses were performed. The cow with highest odds of metritis was a first or greater than or equal to third parity cow, of large breed, that calved during November-April, in a zero-grazing herd. The cow had been treated for dystocia, retained placenta, and at least one other reproductive disease, but not for ketosis.

Estimating transfer of bovine virus-diarrhoea virus in Danish cattle by use of register data.

To study how routinely recorded data (also called "register data") might be used in disease monitoring on a regional or national level, a database for bovine virus-diarrhoea virus (BVDV) was made from existing databases, covering the period January 1995-November 1999. This paper includes a general description of the database, including basic statistics for selected variables. Information was largely complete for cattle herds in the milk-recording scheme (MRS), but only partly available for other herds. A methodology was developed to identify when and how a herd initially was infected. For most herds, it was possible to determine when and how BVDV first was introduced. Among the infected herds, most were already infected by the start of the study. BVDV had been present in 40% of the MRS herds and in 9% of the non-MRS herds. In the MRS herds, most new infections were associated with a dam that had been present in the same herd during gestation. Among the non-MRS herds, most new infections were associated with movement of a persistently infected animal. The monthly number of newly infected herds is presented; it is seen that the incidence declined substantially during the study period.

Welfare in Danish dairy herds 1. Disease management routines in 1983 and 1994.

This paper presents the first part of a questionnaire survey carried out in 2148 Danish dairy herds during 1994, as well as results from a similar survey carried out in 1983. The welfare status in Danish dairy herds with respect to disease management routines currently applied is discussed. In detail this was: recording of mastitis incidents, use of veterinarian for milk fever cases, farmer's effort in reducing incidence of mastitis, milk fever, ketosis, calving problems, and lameness, as well as frequency of claw trimming, reasons for culling, and way of replacing cullings. Furthermore, trends during the 11 year period are discussed. The results show that the Danish dairy farmers in 1994 in general have a substantial knowledge of prevention and treatment of disease. However, adjustments in the following areas would be appropriate: 1) farmers should avoid making intravenous infusions, 2) they should be encouraged to use calving boxes for parturitions, 3) there should be more attention on claw health, and 4) to comply with the new Danish legislation, antibiotic dry cow treatment should only be carried out on the individual cow if pathogenic microorganisms have been isolated within 35 days prior to drying off.

Welfare in Danish dairy herds 2. Housing systems and grazing procedures in 1983 and 1994.

This paper presents the second part of a questionnaire survey carried out in 2148 Danish dairy herds during 1994, as well as results from a similar survey carried out in 1983. The welfare status and trends during the 11 year period are discussed with respect to cattle housing systems and grazing procedures. Generally speaking, the results show that Danish dairy farmers in 1994 followed the common recommendations, i.e. 1) there are partitions between stalls in almost all tie stall houses, 2) feeding cubicles are seldom seen in cubicle houses, 3) bedding is provided for most cows, 4) saw dust as bedding for cows is not commonly used, and 5) the majority of cows and heifers are pastured during summer. However, adjustments in the following areas would be appropriate: 1) tie systems which restrict the cow's natural rising and lying should be phased out, and 2) loose housed heifers in boxes should have access to a bedded resting area.

Welfare in Danish dairy herds 3. Health management and general routines in 1983 and 1994.

This paper presents the third part of descriptive results of questionnaire surveys in 152 Danish dairy herds in 1983 and in 2148 dairy herds in 1994. Focus is on working routines related to health management and the close environment of the cows. The variables are grouped in 6 categories as man power, bedding, water supply, manure handling, health management routines, and the farmers' age and their opinion about health and welfare of the dairy cows. The results show that the husbands did the major parts of the job in the herds. Permanent laborers were mainly hired in cubicle and deep bed farms, while it was more common to hire a relief man (short term basis) in tie stall herds-i.e. in the generally smaller herds. The average time spent on milking and feeding per cow per day ranged from 5.2 min in cubicle houses and 5.4 min in deep bed houses to 9.9 min in tie stall houses. The time per cow per day seemed to have been reduced by approximately 43% during the 11 year period. Straw was the primary choice of bedding, and the use varied much among the herds. In tie stall houses with open dung channel and concrete floor the daily average use of straw was 1.74 kg per cow. Only 37.0% of the farmers used bedding for the heifers. Water supply seemed to be well installed in all houses, and dung removal was highly automated. Apart from milking and feeding times the farmers looked after the cows on average twice a day. The farmers primarily looked for cows in heat, signs of disease, calving, and abnormal lying and raising patterns. At night 87.7% of tie stall farmers and 80.8% of cubicle house farmers were likely to check the cows, particularly with respect to calving. In deep bed systems only 58.5% would check the cows at night. Contrary to this, farmers looked after pastured heifers less frequently. Farmers were generally concerned that the cows had a dry period. The average length stated was 6.6 weeks. Farmers were generally satisfied with the health and welfare of the cows. The answers also indicated that farmers differentiated between the 2 concepts, as the correlation between welfare and health was only r = 0.34 (p < 0.001).

Implantation of temperature loggers in 100 Danish dairy calves: surgical procedure and follow-up.

One hundred Danish dairy calves had temperature loggers implanted subcutaneously on the neck. Post-operatively, the calves were given a single antibiotic treatment, and tissue reactions were assessed on 6 post-operative visits. After approximately 5 months, the loggers were removed and material submitted for histologic examination. This paper presents 1) the surgical procedure, 2) the prevalence of tissue reaction at the post-operative visits, 3) the degree of implant recovery, 4) the results of histopathologic examinations, 5) an evaluation of age at implantation or veterinary practitioner as risk factors for tissue reaction and missing implant recovery 5 months after implantation, and 6) evaluation of tissue reaction as a risk factor for lack of recovery 5 months after implantation. The implant was rejected on 7 calves (7%). Additionally, 5 calves (5%) had the temperature logger removed because of presence of an abcess. No migration of the temperature loggers were observed. The results of a repeated measures analysis and the histopathological findings indicate that contamination during the surgery resulted in inflammation and abcess formation. It is recommended that in the presence of an abcess, the temperature logger should be removed.

Assessment of confidence in freedom from Aujeszky's disease and classical swine fever in Danish pigs based on serological sampling--effect of reducing the number of samples.

Confirming freedom from disease is important for export of animals and animal products. In Denmark, an intensive surveillance program is in place for Aujeszky's disease (AD) and classical swine fever (CSF), including 34,974 blood samples tested for AD and 37,414 samples tested for CSF (2008 figures). In the current system, 3.5% of sows and boars for export or slaughter are tested for both diseases, as well as all boars before entering boar stations. Furthermore, nucleus herds are tested every third month for classical swine fever. We investigated, whether the sample size could be reduced without compromising the posterior probability of freedom (PostPFree) from AD and CSF by use of a scenario tree model. Conventional herds and sows or boars were defined as risk factors (compared to SPF(1) herds and finisher pigs), with a relative risk of 2 and 5, respectively. The probability of introduction was modeled as a distribution (0.0042:0.0083; 0.05), and the within-herd and between-herd design prevalence were set to 0.05 and 0.01, respectively. If 50 and 75% of the test results from exported or slaughtered sows and boars were simulated to be removed at random, while the blood samples from boar stations were kept constant (reflecting a total reduction of 28 or 43%) the PostPFree from AD was reduced from 0.989 after 1 year testing to 0.980 or 0.971, respectively. Similarly, the confidence of freedom from CSF was reduced from 0.989 to 0.982 or 0.969, when the number of serological samples from abattoirs and export sows and boars is reduced by 50 or 75%, respectively (reflecting a total reduction of 34 or 51%), and further to 0.978 or 0.963 if sampling in nucleus herds was stopped (reflecting a total reduction of 41 or 59%). The results show that a reduction in the sampling size of Danish sows will have limited effect on the PostPFree from AD and CSF, and that sampling in nucleus herds for CSF adds little to the PostPFree from CSF.