Generalised linear mixed models analysis of risk factors for contamination of Danish broiler flocks with Salmonella typhimurium

Modeling Salmonella Spread in Broiler Production: Identifying Determinants and Control Strategies

The presence of Salmonella spp. in broiler production is a food safety concern as the bacterium can be transmitted to humans via contaminated meat and derived products. Salmonella detection in litter at the pre-slaughter period has been linked to increased odds of contaminated broiler carcasses and meat derived products. To determine risk factors related to farm and broiler house characteristics and management practices, this study uses a unique longitudinal data set from a Brazilian integrated broiler enterprise, which contains official results of Salmonella spp. isolation from drag swabs collected at the end of the grow-out period. A Bayesian hierarchical spatio-temporal model found significant spatial and time influence on the odds of isolating Salmonella spp. from litter as well as significant effects from the size of a broiler house, total housing area per farm, type of broiler house, and number of litter recycles. Results indicate that recycling litter beyond 6 rearing cycles significantly increased the odds of isolating Salmonella before slaughter, and the bacterium was more likely to persist in conventional broiler houses, compared to broiler houses with controlled environment. Evidence of a potential principal-agent problem was also found in setting strategies to control the bacterium from litter, which suggests strong incentives to adopt the strategies aiming to reduce prevalence of the bacterium in the integrated enterprise. Our findings could be used to develop alternative measures to reduce the risk of persistence of the bacterium in the broiler production chain.

Identification and selection of food safety-related risk factors to be included in the Canadian Food Inspection Agency's Establishment-based Risk Assessment model for Hatcheries

Towards the continuous improvement of its inspection system, the Canadian Food Inspection Agency (CFIA) is developing an Establishment-based Risk Assessment model for Hatcheries to allocate inspection resources based on the food safety risk associated with the Canadian hatcheries falling under its jurisdiction. The objectives of the current study were to identify and select critical food safety-related risk factors that would be included in this model, with a main focus on Salmonella. A literature review was used to develop a comprehensive list of risk factors that could potentially contribute to the food safety risk attributed to Canadian hatcheries operating in all production streams (breeders, layers, broilers, turkeys, waterfowl and game birds). The development of this list used a selection process that was conducted according to the availability of data sources, the clarity of definition and the measurability of the selected risk factors. A panel of experts reviewed and adjusted the identified risk factors. A final list of 29 risk factors was generated; 20 originated from the scientific literature and nine from the expert panel. Risk factors were grouped in three clusters according to whether they pertained to the inherent risk (nine factors identified), risk mitigation (nine factors identified) or compliance of a hatchery with its preventive control plan and regulatory requirements (11 factors identified). Criteria for assessing each risk factor were defined based on common practices used in the Canadian hatchery industry. This comprehensive list of risk factors and criteria represents useful information to support the design and implementation of a Canadian risk assessment model for hatcheries, but could also be used by like-minded food safety authorities.

Salmonella control in poultry flocks and its public health impact

An increase in confirmed human salmonellosis cases in the EU after 2014 triggered investigation of contributory factors and control options in poultry production. Reconsideration of the five current target serovars for breeding hens showed that there is justification for retaining Salmonella Enteritidis, Salmonella Typhimurium (including monophasic variants) and Salmonella Infantis, while Salmonella Virchow and Salmonella Hadar could be replaced by Salmonella Kentucky and either Salmonella Heidelberg, Salmonella Thompson or a variable serovar in national prevalence targets. However, a target that incorporates all serovars is expected to be more effective as the most relevant serovars in breeding flocks vary between Member State (MS) and over time. Achievement of a 1% target for the current target serovars in laying hen flocks is estimated to be reduced by 254,400 CrI95[98,540; 602,700] compared to the situation in 2016. This translates to a reduction of 53.4% CrI95[39.1; 65.7] considering the layer-associated human salmonellosis true cases and 6.2% considering the overall human salmonellosis true cases in the 23 MSs included in attribution modelling. A review of risk factors for Salmonella in laying hens revealed that overall evidence points to a lower occurrence in non-cage compared to cage systems. A conclusion on the effect of outdoor access or impact of the shift from conventional to enriched cages could not be reached. A similar review for broiler chickens concluded that the evidence that outdoor access affects the occurrence of Salmonella is inconclusive. There is conclusive evidence that an increased stocking density, larger farms and stress result in increased occurrence, persistence and spread of Salmonella in laying hen flocks. Based on scientific evidence, an impact of Salmonella control programmes, apart from general hygiene procedures, on the prevalence of Campylobacter in broiler flocks at the holding and on broiler meat at the end of the slaughter process is not expected.

Salmonella spp. transmission in a vertically integrated poultry operation: Clustering and diversity analysis using phenotyping (serotyping, phage typing) and genotyping (MLVA)

The transmission of Salmonella enterica within a vertically integrated poultry operation was investigated longitudinally over an 18-month period (2013–2014). Thirty six percent of all samples collected (1503 of 4219) were positive for salmonellae with seven Salmonella enterica subsp. enterica serovars, and one Salmonella enterica subsp. salamae serovar detected. Both Salmonella enterica subsp. enterica serovars Infantis and Typhimurium were detected in all locations sampled. Salmonella Typhimurium was the most frequently detected serovar (63% of serotyped samples) with 8 phage types (PT) and 41 multiple-locus variable-number tandem-repeats analysis (MLVA) profiles identified. The most frequently identified phage types were PT135a and DT135. A total of 62 PT/MLVA combinations were observed. MLVA profiles 03-14-10-09-525 and 03-15-11-11-525 were the most frequently identified and 83% of the isolates shared at least one MLVA profile with an isolate from another phage type. The use of phage typing and MLVA profiling, on their own or in combination, were insufficient to understand the complexity of the epidemiological relationships between locations within this production system. Despite the high level of apparent diversity, cluster analysis was unable to differentiate the transmission pathways of all S. Typhimurium variants detected within the integrated enterprise. Using additional epidemiological information, the parent breeder rearing site was identified as the most likely point of introduction of two S. Typhimurium isolates into the production system with subsequent dissemination to the broiler flocks via the hatchery. This complexity is unable to be resolved in the absence of intensive sampling programs at all generations of the production system.

Investigation of potential risk factors associated with Salmonella presence in commercial laying hen farms in Nigeria

BACKGROUND

In 2012/13, the Federal government of Nigeria approved the surveillance of salmonellae in commercial poultry farms with the aim of generating baseline data for the development of a control programme. That baseline provided an opportunity to investigate potential risk factors in commercial layer farms.

METHODS

Five hundred and twenty-three farms were evaluated for the presence of Salmonella. Each farmer was interviewed through a structured questionnaire. Univariate analysis identified 26 variables potentially associated with Salmonella presence on the farms, with different levels of significance. To simultaneously evaluate the effect of these variables, a multivariable logistic regression model was implemented.

RESULTS

Farmers that are member of the Poultry Association of Nigeria (PAN) emerged as a risk factor, as well as being farmer as exclusive occupation. The risk of Salmonella increased in farms with a combined housing system as compared to farms with a single housing system. Moreover, in terms of farm management the use of plastic egg crates had a higher risk of Salmonella presence compared to paper crates. Higher risk of contamination was also identified in farms having regular contact with animal health workers, since this could be indicative of animal health problems requiring continuous assistance. Farms with no previous outbreak of salmonellosis were associated with an increase risk, since most of the farms in this category were newly established. With regard to feeding, the use of coccidiostats and commercial feeds were associated with the spread of Salmonella. Unexpectedly, the presence of other farms at distances further than 1 km away was associated with an increase in Salmonella presence as compared to farms which were in closer proximity. Close proximity among farms could favour the application of stricter biosecurity measures.

CONCLUSIONS

The results of the survey can be considered a baseline for implementing effective measures aimed at reducing the Salmonella prevalence in Nigerian laying hen farms.

Three-Level Mixed-Effects Logistic Regression Analysis Reveals Complex Epidemiology of Swine Rotaviruses in Diagnostic Samples from North America

Rotaviruses (RV) are important causes of diarrhea in animals, especially in domestic animals. Of the 9 RV species, rotavirus A, B, and C (RVA, RVB, and RVC, respectively) had been established as important causes of diarrhea in pigs. The Minnesota Veterinary Diagnostic Laboratory receives swine stool samples from North America to determine the etiologic agents of disease. Between November 2009 and October 2011, 7,508 samples from pigs with diarrhea were submitted to determine if enteric pathogens, including RV, were present in the samples. All samples were tested for RVA, RVB, and RVC by real time RT-PCR. The majority of the samples (82%) were positive for RVA, RVB, and/or RVC. To better understand the risk factors associated with RV infections in swine diagnostic samples, three-level mixed-effects logistic regression models (3L-MLMs) were used to estimate associations among RV species, age, and geographical variability within the major swine production regions in North America. The conditional odds ratios (cORs) for RVA and RVB detection were lower for 1–3 day old pigs when compared to any other age group. However, the cOR of RVC detection in 1–3 day old pigs was significantly higher (p < 0.001) than pigs in the 4–20 days old and >55 day old age groups. Furthermore, pigs in the 21–55 day old age group had statistically higher cORs of RV co-detection compared to 1–3 day old pigs (p < 0.001). The 3L-MLMs indicated that RV status was more similar within states than among states or within each region. Our results indicated that 3L-MLMs are a powerful and adaptable tool to handle and analyze large-hierarchical datasets. In addition, our results indicated that, overall, swine RV epidemiology is complex, and RV species are associated with different age groups and vary by regions in North America.

Ecological determinants of highly pathogenic avian influenza (H5N1) outbreaks in Bangladesh

Background

The agro-ecology and poultry husbandry of the south Asian and south-east Asian countries share common features, however, with noticeable differences. Hence, the ecological determinants associated with risk of highly pathogenic avian influenza (HPAI-H5N1) outbreaks are expected to differ between Bangladesh and e.g., Thailand and Vietnam. The primary aim of the current study was to establish ecological determinants associated with the risk of HPAI-H5N1 outbreaks at subdistrict level in Bangladesh. The secondary aim was to explore the performance of two different statistical modeling approaches for unmeasured spatially correlated variation.

Methodology/Principal Findings

An ecological study at subdistrict level in Bangladesh was performed with 138 subdistricts with HPAI-H5N1 outbreaks during 2007–2008, and 326 subdistricts with no outbreaks. The association between ecological determinants and HPAI-H5N1 outbreaks was examined using a generalized linear mixed model. Spatial clustering of the ecological data was modeled using 1) an intrinsic conditional autoregressive (ICAR) model at subdistrict level considering their first order neighbors, and 2) a multilevel (ML) model with subdistricts nested within districts. Ecological determinants significantly associated with risk of HPAI-H5N1 outbreaks at subdistrict level were migratory birds' staging areas, river network, household density, literacy rate, poultry density, live bird markets, and highway network. Predictive risk maps were derived based on the resulting models. The resulting models indicate that the ML model absorbed some of the covariate effect of the ICAR model because of the neighbor structure implied in the two different models.

Conclusions/Significance

The study identified a new set of ecological determinants related to river networks, migratory birds' staging areas and literacy rate in addition to already known risk factors, and clarified that the generalized concept of free grazing duck and duck-rice cultivation interacted ecology are not significant determinants for Bangladesh. These findings will refine current understanding of the HPAI-H5N1 epidemiology in Bangladesh.

Multilevel analysis of environmental Salmonella prevalences and management practices on 49 broiler breeder farms in four south-eastern States, USA

A two-part serial survey of 49 broiler breeder farms was conducted in four south-eastern states: Arkansas, Alabama, Georgia and North Carolina. Broiler breeder farms from three to five broiler company complexes in each state were visited on two separate occasions to document management practices and perform environmental sampling for Salmonella prevalence estimation. Salmonella was detected in 88% of the broiler breeder houses that were sampled and was identified on all 49 farms enrolled. Many management characteristics were consistent across the different states and companies. Multilevel analysis was used to evaluate management characteristics as risk factors for Salmonella prevalence and to estimate the proportion of variance residing at the different hierarchical sampling levels. Management characteristics associated with increased Salmonella prevalence included treatment of the flock for any disease, having dusty conditions in the house, having dry conditions under the slats and walking through the house more than one time per day to pick-up dead birds. After adjusting for state as a fixed effect, the percentages of variance in Salmonella prevalence occurring at the complex, farm, visit, house and individual sample levels were 5.2%, 6.8%, 11.8%, 2.8% and 73.4%, respectively. The intraclass correlations for samples collected from the same house; for samples from different houses during the same visit; for samples from different visits to the same farm; and for samples from different farms in the same complex were as follows: 0.27, 0.24, 0.12 and 0.05, respectively.

Risk factors associated with Salmonella status of broiler flocks delivered to grow-out farms

In a prospective field observational study in the southeastern USA, we sampled gastrointestinal (GI) tracts from chicks of 65 broiler flocks delivered to conventional grow-out farms for rearing. The flocks were hatched at seven broiler hatcheries. The mean within-flock prevalence of Salmonella-positive samples was 6.5% and ranged from 0% to 86.7%. Of the 65 flocks studied, 25 (38.5%) had at least one Salmonella-positive sample. Accounting for confounding variability among the hatcheries and broiler companies, we tested whether the probability of detecting Salmonella in GI tracts of the chicks delivered was associated with certain characteristics of parent breeder flocks; hatchery production volume; hatchery ventilation system; hatchery egg-room conditions; egg incubation, candling, hatching, eggshell and bird separation, and bird-processing procedures; management of hatchery-to-farm transportation; day of week of hatch; weather conditions during transportation; or season of the hatch. Two risk factor models were adopted. The first model indicated that a greater number of parent flocks, manual separation of eggshell and bird, and a greater amount of fluff and feces on tray liners used during hatchery-to-farm transportation at delivery were associated with increased probability of detecting Salmonella in chick GI tracts, whereas a greater number of birds in the delivery vehicle was associated with decreased probability. The second model indicated that broiler flocks hatched on Tuesdays versus either Mondays or Thursdays (with no hatches on Wednesdays, Fridays or week-ends), increased average hatchability of the eggs from the parent flocks, and greater amounts of fluff and feces on the transport tray liners at delivery were all associated with increased probability of detecting Salmonella in chick GI tracts. The results of this study suggest potential management decisions to lessen Salmonella contamination of broilers supplied by commercial hatcheries and areas for further research.

Associations between vaccinations against protozoal and viral infections and Salmonella in broiler flocks

Reducing the burden of Salmonella in broiler flocks presents a challenge for public health. Worldwide, grow-out broilers are routinely vaccinated to prevent or lessen clinical manifestation of other infections. In this exploratory analysis we tested if details of a routine vaccination programme delivered to conventional grow-out broilers were associated with the burden of Salmonella in the flock as it progressed through its production cycle. None of the flocks studied were vaccinated against Salmonella or received a competitive exclusion product. The flocks were reared on conventional grow-out farms in southeastern USA, and sampled in a prospective field observational study. We observed significant associations between the content and design of a grow-out vaccination programme targeting other infections and the probability of detecting Salmonella in the broiler flock at different time points throughout the production cycle. To the best of the authors' knowledge, this is the first field report of such associations.

Investigation of risk factors for Salmonella on fattening-turkey farms

A cross-sectional study into risk factors for Salmonella was undertaken using data gathered from 252 fattening turkey flocks in the UK. The data was derived from the EU baseline survey conducted during 2006 and 2007, in addition to a voluntary questionnaire. Multivariate logistic regression models identified significant risk factors for Salmonella spp. and Salmonella Typhimurium. A decreased risk of Salmonella spp. infection was associated with a history of intestinal illness in the sampled flock (OR 0·17), the use of wood shavings as litter (OR 0·21), use of disinfectant in the cleaning process (OR 0·25), incineration of dead birds on farm (OR 0·29), seasonal production (OR 0·31), farm staff also working with cattle (OR 0·31), and the presence of pigs on neighbouring farms (OR 0·38). The risk of isolating Salmonella spp. varied according to the company from which the poults were sourced. A reduced risk of S. Typhimurium infection was associated with the use of wax blocks to control rodents (OR 0·09), using mains water (OR 0·19) and having a Salmonella test programme (OR 0·23). An increased risk of S. Typhimurium infection was associated with storage of items around the turkey house (OR 5·20), evidence of mice (OR 4·71) and a soil surface surrounding the turkey house (OR 2·70). This study therefore identifies a number of important practical measures which can be implemented by farmers and veterinarians within the turkey industry to assist in the control of salmonellosis at the farm level.

Survey of the prevalence of Salmonella on commercial broiler farms in the United Kingdom, 2005/06

Between October 2005 and September 2006, all European Union member states were required to carry out standardised surveys of the prevalence of Salmonella in broiler flock holdings to establish baseline data from which to derive national targets for disease reduction. In the uk 382 holdings were sampled, 41 of which were positive for Salmonella, giving an estimated weighted prevalence of 10.7 per cent (95 per cent confidence interval [ci] 8.1 to 13.1 per cent). The serotype most frequently isolated was Salmonella Ohio, with a weighted prevalence of 2.2 per cent (95 per cent ci 1.2 to 3.7 per cent), followed by Salmonella Kedougou at 1.7 per cent (95 per cent ci 0.9 to 3.2 per cent). There were no isolates of Salmonella Enteritidis and only a single isolation of Salmonella Typhimurium (0.2 per cent, 95 per cent ci 0.0 to 1.6 per cent). Of the three other serotypes given top priority by the eu owing to their public health significance, Salmonella Virchow was isolated from one holding, but Salmonella Hadar and Salmonella Infantis were not detected on any of the holdings.

Prevalence and risk factors for Salmonella spp. and Campylobacter spp. caecal colonization in broiler chicken and turkey flocks slaughtered in Quebec, Canada

We conducted an observational study to estimate prevalence and risk factors for Salmonella spp. and Campylobacter spp. caecal colonization in poultry. Eighty-one broiler chicken and 59 turkey flocks selected among flocks slaughtered in the province of Quebec, Canada, were included in the study. Flock status was evaluated by culturing pooled caecal contents from about 30 birds per flock. Exposure to potential risk factors was evaluated with a questionnaire. Odds ratios were computed using multivariable logistic regression. The prevalence of Salmonella-positive flocks was 50% (95% CI: 37, 64) for chickens and 54% (95% CI: 39, 70) for turkeys, respectively. Odds of Salmonella colonization were 2.6 times greater for chicken flocks which failed to lock the chicken house permanently. In turkeys, odds of Salmonella colonization were 4.8-7.7 times greater for flocks which failed to be raised by <or=2 producers with no other visitors allowed onto the premises, or origin from a hatchery. The prevalence of Campylobacter-positive flocks was 35% (95% CI: 22, 49) for chickens and 46% (95% CI: 30, 62) for turkeys. Odds of colonization were 4.1 times higher for chicken flocks raised on farms with professional rodent control and 5.2 times higher for flocks with manure heap >200m from the poultry house, and also increased with the number of birds raised per year on the farm and with the age at slaughter. For turkeys, odds of Campylobacter flock colonization were 3.2 times greater in flocks having a manure heap at </=200m from poultry house and 4.2 times greater in flocks drinking unchlorinated water.

An epidemic of salmonellosis caused by Salmonella Typhimurium DT160 in wild birds and humans in New Zealand

AIM

This study reports an outbreak of salmonellosis due to S. Typhimurium DT160 which caused extensive mortality in wild birds and enteric disease in humans in New Zealand during the winter and spring months of the year 2000.

METHODS

Necropsies were performed and microbiological examinations undertaken on wild birds from populations in which mass mortality was reported, and on captive indigenous birds which died suddenly during the winter and spring of 2000. Affected tissues were examined histologically and isolates of S. Typhimurium were phage typed and examined using pulsedfield gel electrophoresis (PFGE). Isolates of S. Typhimurium obtained from cases of human enteric disease which occurred during these months were phage typed, examined using PFGE and compared with the bird isolates.

RESULTS

Central and northern areas of the South Island and the southern North Island were worst affected with die-offs of several hundreds of sparrows and other birds reported in rural areas. Mortalities reached a peak in winter (July-August) 2000 and decreased to small numbers during the spring and early summer. The birds usually died of an acute septicaemia with multifocal necrotising lesions in the liver and spleen. Human cases throughout the country increased gradually over the same period. Isolates from birds, livestock and humans examined using PFGE were indistinguishable from one another.

CONCLUSION

This strain of Salmonellahas emerged as a major cause of septicaemia in wild birds in New Zealand. Because of the close association between house sparrows (Passer domesticus) and humans, the organism also poses a serious zoonotic risk. The possibility that the infection may spread to involve indigenous species needs investigation.

Risk factors for Salmonella enterica subsp. enterica infection in senegalese broiler-chicken flocks

Our objective was to assess the association of managerial practices, general hygiene and Salmonella infection in Senegalese broiler flocks. Seventy broilers farms were studied from January 2000 to December 2001 around Dakar. A questionnaire was submitted to the farmers and samples of fresh broiler droppings were taken. A 28.6% of the flocks were infected by Salmonella (mainly Hadar and Brancaster serovars). Salmonella infection of the previous flock (OR = 6.82) and of day-old chicks (OR = 3.73), frequent poultry farmers' visits (OR = 5.38) and keeping sick birds inside the farm (OR = 5.32) increased the risk of Salmonella infection. But, using antibiotics on day-old chicks (OR = 0.17) and a detergent for cleaning (OR = 0.16) decreased the risk.

Estimation of infection prevalence from correlated binomial samples

Infection prevalence in a population often is estimated from grouped binary data expressed as proportions. The groups can be families, herds, flocks, farms, etc. The observed number of cases generally is assumed to have a Binomial distribution and the estimate of prevalence is then the sample proportion of cases. However, the individual binary observations might not be independent--leading to overdispersion. The goal of this paper was to demonstrate random-effects models for the estimation of infection prevalence from data which are correlated and in particular, to illustrate a nonparametric random-effects model for this purpose. The nonparametric approach is a relatively recent addition to the random-effects class of models and does not appear to have been discussed previously in the veterinary epidemiology literature. The assumptions for a logistic-regression model with a nonparametric random effect were outlined. In a demonstration of the method on data relating to Salmonella infection in Irish pig herds, the nonparametric method resulted in the classification of herds into a small number of distinct prevalence groups (i.e. low, medium and high prevalence) and also estimated the relative frequency of each prevalence category in the population. We compared the estimates from a logistic model with a nonparametric distribution for the random effects with four alternative models: a logistic-regression model with no random effects, a marginal model using a generalised estimating equation (GEE) and two methods of fitting a Normally distributed random effect (the GLIMMIX macro and the NLMIXED procedure both in SAS). Parameter estimates from random-effects models are not readily interpretable in terms of prevalences. Therefore, we outlined two methods for calculating population-averaged estimates of prevalence from random-effects models: one using numerical integration and the other using Monte Carlo simulation.

Sources of variation in milk urea nitrogen in Ohio dairy herds

The purpose of this study was to estimate the amount of variation in milk urea nitrogen (MUN) concentrations attributable to test-day, individual cow, and herd effects and to describe factors associated with MUN measurements in Ohio dairy herds. The data came from 24 Holstein herds, half of which were classified as low producing (LP) [rolling herd average (RHA) milk production < 7,258 kg] and half as high producing (HP) herds (RHA production > 10,433 kg). MUN concentration was measured from cow's monthly test-day milk samples. The data were analyzed using multilevel modeling technique in MLwiN, separately for LP and HP herds. The unadjusted mean MUN was 13.9 mg/dl for the HP herds and 11.3 mg/dl for the LP herds. The variance structure was different between the two groups. Most of the variability was found at test-day level in the LP herds, but at herd level in HP herds. MUN was lowest during the first month of lactation, and also season was associated with MUN in both groups. Test-day milk yield, milk fat percentage, and SCC were associated with MUN in the HP herds. With significant explanatory variables in the model, proportionally more of the variation was explained at herd level and less at test day level in both groups. Lower variability in MUN between test days in the HP herds may indicate more consistent day-to-day feeding and management within a herd. The great variability between test days should be considered when interpreting MUN and samples should be collected at the same time of the day to minimize day-to-day variability.

A probabilistic transmission model of Salmonella in the primary broiler production chain

Annual data from the Finnish National Salmonella Control Programme were used to build up a probabilistic transmission model of salmonella in the primary broiler production chain. The data set consisted of information on grandparent, parent, and broiler flock populations. A probabilistic model was developed to describe the unknown true prevalences, vertical and horizontal transmissions, as well as the dynamical model of infections. By combining these with the observed data, the posterior probability distributions of the unknown parameters and variables could be derived. Predictive distributions were derived for the true number of infected broiler flocks under the adopted intervention scheme and these were compared with the predictions under no intervention. With the model, the effect of the intervention used in the programme, i.e., eliminating salmonella positive breeding flocks, could be quantitatively assessed. The 95% probability interval of the posterior predictive distribution for (broiler) flock prevalence under current (1999) situation was [1.3%-17.4%] (no intervention), and [0.9%-5.8%] (with intervention). In the scenario of one infected grandparent flock, these were [2.8%-43.1%] and [1.0%-5.9%], respectively. Computations were performed using WinBUGS and Matlab softwares.

Observations on the distribution and control of Salmonella species in two integrated broiler companies

The effectiveness of cleaning and disinfecting broiler farms and the persistence of Salmonella species in two integrated broiler companies was investigated for two years. Both companies used a cleaning and disinfection regime which included the application of a spray of phenolic disinfectant followed by fogging with formaldehyde solution, and this was highly effective in preventing carry-over of infection in the broiler houses. The disinfection of service areas and areas outside the houses was less effective but it had no influence on the Salmonella status of later flocks. Both companies had persistent problems with the contamination of pellet cooling systems in their feedmills with Salmonella 4, 12:d:- in company A, and with Salmonella binza and Salmonella ohio in company B. The hatcher incubators of both companies were also persistently contaminated with Salmonella livingstone and Salmonella thomasville in company A and with Salmonella senftenberg in company B. At both companies sites Salmonella enteritidis and Salmonella typhimurium Tr104 were also isolated occasionally from various locations.

The use of multilevel models to evaluate sources of variation in reproductive performance in dairy cattle in Reunion Island

Sources of variation in measures of reproductive performance in dairy cattle were evaluated using data collected from 3207 lactations in 1570 cows in 50 herds from five geographic regions of Reunion Island (located off the east coast of Madagascar). Three continuously distributed reproductive parameters (intervals from calving-to-conception, calving-to-first-service and first-service-to-conception) were considered, along with one Binomial outcome (first-service-conception risk). Multilevel models which take into account the hierarchical nature of the data were used to fit all models. For the overall measure of calving-to-conception interval, 86% of the variation resided at the lactation level with only 7, 6 and 2% at the cow, herd and regional levels, respectively. The proportion of variance at the herd and cow levels were slightly higher for the calving-to-first-service interval (12 and 9%, respectively) - but for the other two parameters (first-service-conception risk and first-service-to-conception interval), >90% of the variation resided at the lactation level. For the three continuous dependent variables, comparison of results between models based on log-transformed data and Box-Cox-transformed data suggested that minor departures from the assumption of normality did not have a substantial effect on the variance estimates. For the Binomial dependent variable, five different estimation procedures (penalised quasi-likelihood, Markov-Chain Monte Carlo, parametric and non-parametric bootstrap estimates and maximum-likelihood) yielded substantially different results for the estimate of the cow-level variance.

A hierarchical trend model for bovine virus diarrhoea virus (BVDV) sero-conversion in Norwegian dairy herds from 1993 through 1997

The Norwegian bovine virus diarrhoea (BVD) control-and-eradication program performed annual testing of the national cattle population since its first screening in January 1993. A bulk-tank milk ELISA antibody test was used for the initial screening of dairy herds. Based on the annual bulk-tank milk-test result, a binary variable denoting herd sero-conversion (a surrogate measure for incidence) was created. The count of herds with sero-conversion in each community was regressed on year and the initial herd-sero-prevalence for each community in a Poisson hierarchical trend model - modelling the risk of sero-conversion. By using this modelling approach, estimates of trend specific for each hierarchical level of organisation were included in the trend model (community, veterinary district and county) could be estimated. The main BVD trend showed a steadily declining sero-conversion risk. The communities in the highest herd-sero-prevalence quartile in 1993 continued to have the highest sero-conversion risk throughout the study period--decreasing from an average predicted sero-conversion risk in 1993 of 0.12 (95% CI; 0.10, 0.13) to 0.02 (0.007, 0.04) in 1997. There was an expressed variation in the level of sero-conversion for all the three hierarchical levels, but the trend only varied at the lowest level (community).