One-Health Simulation Modelling: Assessment of Control Strategies Against the Spread of Influenza between Swine and Human Populations Using NAADSM.

Simulation models implemented using a range of parameters offer a useful approach to identifying effective disease intervention strategies. The objective of this study was to investigate the effects of key control strategies to mitigate the simultaneous spread of influenza among and between swine and human populations. We used the pandemic H1N1 2009 virus as a case study. The study population included swine herds (488 herds) and households-of-people (29,707 households) within a county in Ontario, Canada. Households were categorized as: (i) rural households with swine workers, (ii) rural households without swine workers and (iii) urban households without swine workers. Seventy-two scenarios were investigated based on a combination of the parameters of speed of detection and control strategies, such as quarantine strategy, effectiveness of movement restriction and ring vaccination strategy, all assessed at three levels of transmissibility of the virus at the swine-human interface. Results showed that the speed of detection of the infected units combined with the quarantine strategy had the largest impact on the duration and size of outbreaks. A combination of fast to moderate speed of the detection (where infected units were detected within 5-10 days since first infection) and quarantine of the detected units alone contained the outbreak within the swine population in most of the simulated outbreaks. Ring vaccination had no added beneficial effect. In conclusion, our study suggests that the early detection (and therefore effective surveillance) and effective quarantine had the largest impact in the control of the influenza spread, consistent with earlier studies. To our knowledge, no study had previously assessed the impact of the combination of different intervention strategies involving the simultaneous spread of influenza between swine and human populations.

One-Health Simulation Modelling: A Case Study of Influenza Spread between Human and Swine Populations using NAADSM.

The circulation of zoonotic influenza A viruses including pH1N1 2009 and H5N1 continue to present a constant threat to animal and human populations. Recently, an H3N2 variant spread from pigs to humans and between humans in limited numbers. Accordingly, this research investigated a range of scenarios of the transmission dynamics of pH1N1 2009 virus at the swine-human interface while accounting for different percentages of swine workers initially immune. Furthermore, the feasibility of using NAADSM (North American Animal Disease Spread Model) applied as a one-health simulation model was assessed. The study population included 488 swine herds and 29, 707 households of people within a county in Ontario, Canada. Households were categorized as follows: (i) rural households with swine workers, (ii) rural households without swine workers, and (iii) urban households without swine workers. Forty-eight scenarios were investigated, based on the combination of six scenarios around the transmissibility of the virus at the interface and four vaccination coverage levels of swine workers (0-60%), all under two settings of either swine or human origin of the virus. Outcomes were assessed in terms of stochastic 'die-out' fraction, size and time to peak epidemic day, overall size and duration of the outbreaks. The modelled outcomes indicated that minimizing influenza transmissibility at the interface and targeted vaccination of swine workers had significant beneficial effects. Our results indicate that NAADSM can be used as a framework to model the spread and control of contagious zoonotic diseases among animal and human populations, under certain simplifying assumptions. Further evaluation of the model is required. In addition to these specific findings, this study serves as a benchmark that can provide useful input to a future one-health influenza modelling studies. Some pertinent information gaps were also identified. Enhanced surveillance and the collection of high-quality information for more accurate parameterization of such models are encouraged.

Network analysis of swine shipments in Ontario, Canada, to support disease spread modelling and risk-based disease management.

Understanding contact networks are important for modelling and managing the spread and control of communicable diseases in populations. This study characterizes the swine shipment network of a multi-site production system in southwestern Ontario, Canada. Data were extracted from a company's database listing swine shipments among 251 swine farms, including 20 sow, 69 nursery and 162 finishing farms, for the 2-year period of 2006 to 2007. Several network metrics were generated. The number of shipments per week between pairs of farms ranged from 1 to 6. The medians (and ranges) of out-degree were: sow 6 (1-21), nursery 8 (0-25), and finishing 0 (0-4), over the entire 2-year study period. Corresponding estimates for in-degree of nursery and finishing farms were 3 (0-9) and 3 (0-12) respectively. Outgoing and incoming infection chains (OIC and IIC), were also measured. The medians (ranges) of the monthly OIC and IIC were 0 (0-8) and 0 (0-6), respectively, with very similar measures observed for 2-week intervals. Nursery farms exhibited high measures of centrality. This indicates that they pose greater risks of disease spread in the network. Therefore, they should be given a high priority for disease prevention and control measures affecting all age groups alike. The network demonstrated scale-free and small-world topologies as observed in other livestock shipment studies. This heterogeneity in contacts among farm types and network topologies should be incorporated in simulation models to improve their validity. In conclusion, this study provided useful epidemiological information and parameters for the control and modelling of disease spread among swine farms, for the first time from Ontario, Canada.

Burden and cost of gastroenteritis in a Canadian community.

This study estimated the health burden and costs associated with gastroenteritis in the City of Hamilton (Ontario, Canada). The number of cases, number of different resource units used, and cost per resource unit were represented by probability distributions and point estimates. These were subsequently integrated in a stochastic model to estimate the overall burden and cost in the population and to depict the uncertainty of the estimates. The estimated mean annual cost per capita was Can dollar 115. The estimated mean annual cost per case was Can dollar 1,089 and was similar to other published figures. Gastroenteritis represented a significant burden in the study population, with costs high enough to justify prevention efforts. These results, currently the most accurate available estimates for a Canadian population, can inform future economic evaluations to determine the most cost effective measures for reducing the burden and cost of gastroenteritis in the community.

Magnitude and distribution of acute, self-reported gastrointestinal illness in a Canadian community.

To estimate the magnitude and distribution of self-reported, acute gastrointestinal illness in a Canadian-based population, we conducted a retrospective, cross-sectional telephone survey of approximately 3500 randomly selected residents of the city of Hamilton (Ontario, Canada) from February 2001 to February 2002. The observed monthly prevalence was 10% (95 % CI 9.94-10.14) and the incidence rate was 1.3 (95 % CI 1.1-1.4) episodes per person-year; this is within the range of estimates from other developed countries. The prevalence was higher in females and in those aged < 10 years and 20-24 years. Overall, prevalence peaked in April and October, but a different temporal distribution was observed for those aged < 10 years. Although these data were derived from one community, they demonstrate that the epidemiology of acute gastrointestinal illness in a Canadian-based population is similar to that reported for other developed countries.

Evaluation of the Reveal and SafePath rapid Escherichia coli O157 detection tests for use on bovine feces and carcasses.

The Reveal (Neogen Corp., Lansing, Mich.) and SafePath (SafePath Laboratories LLC, St. Paul, Minn.) tests were evaluated for their performance as beef fecal and beef carcass Escherichia coli O157:H7 monitoring tests. Agreement between these tests and a reference test system was determined using naturally contaminated bovine feces and beef carcasses. The reference system utilized immunomagnetic separation with plating onto cefixime, tellurite, sorbitol MacConkey agar, followed by colony testing using a serum agglutination test for the O157 antigen. Relative to this reference method, the Reveal test showed a sensitivity of 46% and a specificity of 82% on bovine feces and a specificity of 99% on carcass samples. The SafePath test, demonstrated a significantly higher sensitivity at 79% and a similar specificity of 79%. On carcass samples the SafePath test performed similarly to the Reveal test, demonstrating a specificity of 100% relative to the reference system. There was an insufficient number of E. coli O157-positive carcass samples to estimate precisely the sensitivity of these two methods. Both methods show promise as rapid carcass monitoring tests, but further field testing to estimate sensitivity is needed to complete their evaluation. The proportion of positive fecal samples for E. coli O157:H7 by the reference method ranged from 10.2% to 36% in 10 lots of cattle with an overall mean of 17.3% (39/225). Quarter carcass sponging of 125 carcasses revealed 1.6% positive for the pathogen (2/125).

Estimates of within-herd incidence rates of Mycobacterium bovis in Canadian cattle and cervids between 1985 and 1994.

We analysed the individual-animal data from six of the nine outbreaks of tuberculosis in Canadian cattle and cervids from 1985 to 1994. A "positive/reactor" animal was one which had either a positive culture or a positive or suspicious reaction on a mid-cervical, comparative cervical, or gross or histopathological test for tuberculosis. Individual-animal data were collected only for herds which had one or more positive/reactor animals. Data were collected from the outbreak records in the Regional or District offices of Agriculture and Agri-food Canada's Animal and Plant Health Directorate. The within-herd spread of Mycobacterium bovis was studied by determining the most-likely date at which the herd was first exposed to M. bovis and the number of reactions which had developed by the time the herd was investigated. The animal-time units at risk in the herd were probably overestimated, resulting in conservative estimates of the within-herd incidence rates. Negative-binomial regression was used to investigate factors which might have influenced the within-herd spread of tuberculosis. Increasing age appeared to be a risk factor for being a positive/reactor animal. When compared to animals 0-12 months old, animals 13-24 months old had an incidence rate ratio (IRR) of 7.6, while animals >24 months old had an IRR of 10.4 (p=0.009). Actual and predicted incidence rates for tuberculosis in mature (>24 months old) animals were calculated. Actual and predicted incidence rates were similar for cervids, within an outbreak. There was more variability between actual and predicted rates in the dairy and beef animals. In the one outbreak (Ontario) where there were positive/reactor cervid, dairy and beef herds, the actual incidence rate for cervids (IR=9.3 cases per 100 animal-years) was almost twice that of dairy cattle (IR=5.0) and three times that of beef cattle (IR=3.1).

Risk factors for the between-herd spread of Mycobacterium bovis in Canadian cattle and cervids between 1985 and 1994.

Microorganisms of the genus Mycobacterium cause tuberculosis in many animal species including humans. Generally, Mycobacterium bovis (M. bovis) infects cattle and cervids, but it has the potential to infect virtually all species of mammals. This study examined and analysed the data from the nine outbreaks of tuberculosis in Canadian cattle and cervids from 1985 to 1994. For the purposes of this study, a positive herd was one with at least one culture-positive animal. A reactor herd had at least one animal which was positive or suspicious on a mid-cervical, comparative cervical, or gross or histopathologic test for tuberculosis. Herd classification was either reactor/positive or negative. Data for the study were collected from the outbreak records in the Regional or District offices of Agriculture and Agri-Food Canada. Logistic regression was used to study spread of tuberculosis between herds. Two risk factors were identified: increasing herd size; and, the reason why a herd was investigated as part of the outbreak. This latter factor was interpreted as a surrogate measure for the nature of contact between the study herd and other potentially infected herds in the outbreak. Increasing herd size was associated with an increased risk of being positive for tuberculosis with herds of 16-35, 36-80, and >80 animals having odds ratios of 2.9, 5.8, and 9.3, respectively, when compared to a herd size of <16 animals (p < 0.001). When compared to perimeter testing (i.e. testing herds within a specified radius of an infected herd), all other reasons for investigation had higher odds ratios (p < 0.001). These odds ratios were 57.8 for traceout herds (i.e. herds which had purchased animal(s) from a reactor/positive herd), 31.8 for herds with pasture or fence-line contact with a reactor/positive herd, and 14.9 for traceback herds (i.e. herds which had been a source of animals for reactor/positive herd(s)).

A simulation model for studying the role of pre-slaughter factors on the exposure of beef carcasses to human microbial hazards.

A Monte Carlo simulation model was constructed for assessing the quantity of microbial hazards deposited on cattle carcasses under different pre-slaughter management regimens. The model permits comparison of industry-wide and abattoir-based mitigation strategies and is suitable for studying pathogens such as Escherichia coli O157:H7 and Salmonella spp. Simulations are based on a hierarchical model structure that mimics important aspects of the cattle population prior to slaughter. Stochastic inputs were included so that uncertainty about important input assumptions (such as prevalence of a human pathogen in the live cattle-population) would be reflected in model output. Control options were built into the model to assess the benefit of having prior knowledge of animal or herd-of-origin pathogen status (obtained from the use of a diagnostic test). Similarly, a facility was included for assessing the benefit of re-ordering the slaughter sequence based on the extent of external faecal contamination. Model outputs were designed to evaluate the performance of an abattoir in a 1-day period and included outcomes such as the proportion of carcasses contaminated with a pathogen, the daily mean and selected percentiles of pathogen counts per carcass, and the position of the first infected animal in the slaughter run. A measure of the time rate of introduction of pathogen into the abattoir was provided by assessing the median, 5th percentile, and 95th percentile cumulative pathogen counts at 10 equidistant points within the slaughter run. Outputs can be graphically displayed as frequency distributions, probability densities, cumulative distributions or x-y plots. The model shows promise as an inexpensive method for evaluating pathogen control strategies such as those forming part of a Hazard Analysis and Critical Control Point (HACCP) system.

Pre-slaughter control of Escherichia coli O157 in beef cattle: a simulation study.

A stochastic simulation model was used to assess the benefit of measures implemented in the pre-slaughter period that are aimed at reducing the contamination of beef carcasses with Shiga-like-toxin-producing Escherichia coli O157. The scenario studied was based on an abattoir processing approximately 1000 head of lot-fed cattle per day. Input assumptions were described using probability distributions to reflect uncertainty in their true values. Control measures that were assessed were based on either a reduction in herd prevalence of infection, reduction in opportunity for cross-contamination in the processing plant by re-ordering of the slaughter queue, reduction of concentration of E. coli O157 in fresh faeces, or a reduction in the amount of faeces, mud and bedding ('tag') transferred from the hide to the carcass. Some control measures evaluated were hypothetical in nature and were included to assist with the planning of research priorities. Simulations suggested that the greatest potential impact is associated with vaccination and with an agent that reduces shedding E. coli O157 in faeces. Knowledge of herd-test results obtained by testing a sample of animals from the herd provides only a minor advantage in control programmes, although application of a rapid test to all animals in all lots might be of some benefit. Under most scenarios, there is ample opportunity for cross-contamination to occur within the slaughter plant as a result of early entry of cattle contaminated with E. coli O157. An industry-wide reduction in the amount of tag attached to hides and addition of a source of cattle having a prolonged average fasting time were not predicted to have a large impact on mean amount of carcass contamination with E. coli O157.

Reliability of an ordinal rating system for assessing the amount of mud and feces (tag) on cattle hides at slaughter.

A study was conducted to provide a quantitative description of the amount of tag (mud, soil, and bedding) adhered to the hides of feedlot beef cattle and to appraise the statistical reliability of a subjective rating system for assessing this trait. Initially, a single rater obtained baseline data by assessing 2,417 cattle for 1 month at an Ontario beef processing plant. Analysis revealed that there was a strong tendency for animals within sale-lots to have a similar total tag score (intralot correlation = 0.42). Baseline data were summarized by fitting a linear model describing an individual's total tag score as the sum of their lot mean tag score (LMTS) plus an amount representing normal variation within the lot. LMTSs predicted by the linear model were adequately described by a beta distribution with parameters nu = 3.12 and omega = 5.82 scaled to fit on the 0-to-9 interval. Five raters, trained in use of the tag scoring system, made 1,334 tag score observations in a commercial abattoir, allowing reliability to be assessed at the individual level and at the lot level. High values for reliability were obtained for individual total tag score (0.84) and lot total tag score (0.83); these values suggest that the tag scoring system could be used in the marketing and slaughter of Ontario beef cattle to improve the cleanliness of animals presented for slaughter in an effort to control the entry of microbial contamination into abattoirs. Implications for the use of the tag scoring system in research are discussed.

Evaluation of accuracy and repeatability of identification of food-borne pathogens by automated bacterial identification systems.

The performances of five automated microbial identification systems, relative to that of a reference identification system, for their ability to accurately and repeatedly identify six common food-borne pathogens were assessed. The systems assessed were the MicroLog system (Biolog Inc., Hayward, Calif.), the Microbial Identification System (MIS; MIDI Inc., Newark, Del.), the VITEK system (bioMérieux Vitek, Hazelwood, Mo.), the MicroScan WalkAway 40 system (Dade-MicroScan International, West Sacramento, Calif.), and the Replianalyzer system (Oxoid Inc., Nepean, Ontario, Canada). The sensitivities and specificities of these systems for the identification of food-borne isolates of Bacillus cereus, Campylobacter jejuni, Listeria monocytogenes, Staphylococcus aureus, Salmonella spp., and verotoxigenic Escherichia coli were determined with 40 reference positive isolates and 40 reference negative isolates for each pathogen. The sensitivities of these systems for the identification of these pathogens ranged from 42.5 to 100%, and the specificities of these systems for the identification of these pathogens ranged from 32.5 to 100%. Some of the systems had difficulty correctly identifying the reference isolates when the results were compared to those from the reference identification tests. The sensitivity of MIS for the identification of S. aureus, B. cereus, E. coli, and C. jejuni, for example, ranged from 47.5 to 72. 5%. The sensitivity of the Microlog system for the identification of E. coli was 72.5%, and the sensitivity of the VITEK system for the identification of B. cereus was 42.5%. The specificities of four of the five systems for the identification of all of the species tested with the available databases were greater than or equal to 97.5%; the exception was MIS for the identification of C. jejuni, which displayed a specificity of 32.5% when it was tested with reference negative isolates including Campylobacter coli and other Campylobacter species. All systems had >80% sensitivities for the identification of Salmonella species and Listeria species at the genus level. The repeatability of these systems for the identification of test isolates ranged from 30 to 100%. Not all systems included all six pathogens in their databases; thus, some species could not be tested with all systems. The choice of automated microbial identification system for the identification of a food-borne pathogen would depend on the availability of identification libraries within the systems and the performance of the systems for the identification of the pathogen.

Validation of the LacTek test applied to spiked extracts of tissue samples: determination of performance characteristics.

LacTek tests are competitive enzyme-linked immunosorbent assays intended for rapid detection of antimicrobial residues in bovine milk. In this study, the LacTek test protocol was modified for use with extracts of bovine tissue to detect beta-lactam, tetracycline, and sulfamethazine residues. Test performance characteristics--precision, accuracy, ruggedness, practicability, and analytical specificity and sensitivity--were investigated. Results suggest that LacTek tests can be easily adapted to detect antimicrobial residues in extracts of lean ground beef. However, positive samples may not contain residues at violative concentrations (i.e., Canadian maximum residue limits), and therefore, additional analysis would be required for final confirmation and quantitation (e.g., chromatography).

A general framework illustrating an approach to quantitative microbial food safety risk assessment.

Hazard analysis critical control point (HACCP), risk assessment, predictive microbiology, and dose-response modeling have been recognized as important tools for the assessment and management of health risks posed by food-borne pathogens. Unfortunately, the biology of both the food chain and food poisoning is complex and dynamic. Therefore, mathematical modeling of microbial risk from food production through to consumption and illness is difficult. Nevertheless, previous authors have made impressive progress in modeling specific pathogen-food-consumer combinations. In this study a framework for a Monte Carlo model of a generic food system was developed. It links together food ingredients, batch processing, cross contamination, microbial growth, cooking, recontamination, consumption, human exposure to pathogens, the dose-response relationship, and the biologic and economic impact components of such risks. This framework is presented to illustrate one potential approach to quantitative risk assessment for microbial food safety. It requires refinement with appropriate distributions and mathematical relationships before it can be applied to a specific pathogen-food-consumer situation.

Antimicrobial drug residues in milk and meat: causes, concerns, prevalence, regulations, tests, and test performance.

This paper presents a historical review of antimicrobial use in food animals, the causes of residues in meat and milk, the types of residues found, their regulation in Canada, tests used for their detection, and test performance parameters, with an emphasis on immunoassay techniques. The development of residue detection methods began shortly after the introduction of antimicrobials to food animal production in the late 1940s. From initial technical concerns expressed by the dairy industry to the present public health and international trade implications, there has been an ongoing need for reliable, sensitive, and economical methods for the detection of antimicrobial residues in food animal products such as milk and meat. Initially there were microbial growth inhibition tests, followed by more sensitive and specific methods based on receptor binding, immunochemical, and chromatographic principle. An understanding of basic test performance parameters and their implications is essential when choosing an analytical strategy for residue testing. While each test format has its own attributes, none test will meet all the required analytical needs. Therefore the use of a tiered or integrated system employing assays designated for screening and confirmation is necessary to ensure that foods containing violative residues are not introduced into the food chain.

Characteristics of the LacTek test as applied to tissue samples: assessment of performance using incurred field samples.

The Lactek test, marketed for antimicrobial residue detection in milk, was validated for the detection of antimicrobial residues in tissues. A previous study found that the LacTek test could confidently identify tissue samples spiked with antimicrobial residues. However, the test could not reliably distinguish violative from nonviolative spiked samples relative to Canadian maximum residue limits (MRLs). The objectives of this study were to assess and compare the performance of the LacTek tests for beta-lactams, tetracyclines, gentamicin, and sulfamethazine on samples containing naturally incurred residues by running the test in parallel with the standard microbial inhibition test (MIT) presently used for the routine testing of tissues at our facility and to assess the agreement with high pressure liquid chromatographic (HPLC) determinative methods. Parallel testing with the official MIT found that the Lactek tests could be confidently used for testing tissue samples containing incurred residues. Among 1,008 MIT-positive samples, the LacTek test found that 90% contained beta-lactams and/or tetracyclines. A further 7.3% of violative residues could not be identified to an antimicrobial class. In addition, 9% of samples testing negative on the MIT were found to contain an antimicrobial residue by the LacTek tests. Comparative testing with HPLC methods found that there was very good agreement between the two tests and that most violations were due to penicillin G and oxytetracycline. Although the LacTek test cannot be used to distinguish violative from nonviolative residue levels, it does offer several advantages over the present MIT. These include speed, ease of use, the ability to identify residues to a specific class, and an improved sensitivity at the MRL level for the most commonly found antimicrobials in tissue.

Contaminants of non-biological origin in foods from animals.

The authors provide an overview of non-biological contaminants in foods from animals. These contaminants comprise chemical and physical hazards which may be introduced during animal production, slaughter and processing or packaging. Emphasis in this paper is placed on those residues which are of most interest to Veterinary Services and for which Veterinary Services have responsibility, namely: residues of veterinary drugs, industrial chemicals, heavy metals and pesticides which may be introduced during animal production. The most contentious residues which occur in meat, milk and eggs are antibacterial drugs, hormonal growth promoters and certain pesticides, heavy metals and industrial chemicals. While rare incidents of human disease have been attributed to hazardous levels of these contaminants in milk and meat, residues of chemical contaminants in foods of animal origin are, in general, rarely detected at more than trace levels and consequently are not of major public health concern. Nevertheless, non-biological contaminants continue to be very important with respect to international trade and consumer confidence, and efforts to reduce the incidence of occurrence in foods is warranted. Furthermore, continued monitoring and periodic reassessment of risks posed by these contaminants is needed to detect or anticipate new problems so that appropriate action can be taken in the interests of public safety.

Analysis of whole-cell fatty acid profiles of verotoxigenic Escherichia coli and Salmonella enteritidis with the microbial identification system.

Differentiation of strains within bacterial species, based on gas chromatographic analysis of whole-cell fatty acid profiles, was assessed with 115 strains of verotoxigenic Escherichia coli and 315 strains of Salmonella enteritidis. Fatty acid-based subgroups within each of the two species were generated. Variability of fatty acid profiles observed in repeat preparations from the same strain approached that observed between subgroups, limiting the usefulness of using fatty acid profiles to subgroup verotoxigenic E. coli and S. enteritidis strains.

A serological survey of bovine syncytial virus in Ontario: associations with bovine leukemia and immunodeficiency-like viruses, production records, and management practices.

Of the 920 cows tested, 56.7% showed antiretroviral serological reactivity. Prevalence rates (95% confidence interval) of antiretroviral antibodies among individual dairy cows in Ontario were: BIV 5.5% (4.0-7.0), BLV 25.7% (22.9-28.6), and BSV 39.6% (36.4-42.8). The following percentages of cows showed serological reactivity against the specified retroviruses: BIV 2.3%, BLV 14.0%, BSV 27.5%, BIV and BSV 1.3%, BIV and BLV 0.9%, BLV and BSV 9.9%, BIV and BLV and BSV 0.9%. These rates of sero-positivity are similar to those found in other countries. Serological test results were not adjusted for sensitivity and specificity. The prevalence rates of antibodies to the three retroviruses (BIV, BLV, and BSV) were significantly different, but no associations were observed between specific retroviral serological test results among individual cows. The prevalence rates of BIV and BSV seropositivity were constant across Ontario, whereas, there was a significant trend for the prevalence rate of BLV seropositivity to decrease going from southwestern to eastern Ontario; cows in eastern Ontario had approximately half the prevalence rate of those in southwestern Ontario. Cows that were seropositive for BSV were significantly older than BSV seronegative cows. There was no association between culling rate and BSV serology. Significant negative associations were found with winter or summer housing of calves separate from adults and summer outdoor exercise for dry cows. The use of calf hutches in the summer had a significant positive association with BSV seropositivity. Regression analyses were done to assess the association of retroviral (BIV, BLV, and BSV) seropositivity on calving interval, milk somatic cell count, and milk production. Serological test results for BIV, BLV, and BSV were entered into all models and all models were adjusted for intra-cluster (intraherd) correlation. Herd size and age were found to be important confounding variables. BIV seropositivity was not associated with any changes in production using this approach, however when considered in isolation BIV seropositivity remained associated with decreased milk production. BLV seropositivity was significantly associated with longer calving intervals and higher somatic cell counts in older cows. As well, in older cows, BSV seropositivity was significantly associated with higher milk production.

A serological survey for bovine immunodeficiency-like virus in Ontario dairy cattle and associations between test results, production records and management practices.

A chemiluminescence Western blot analysis (WBA) for detecting antibovine immunodeficiency-like virus (BIV) antibodies, had good repeatability. The test was subsequently applied to a bank of serum samples from 928 adult cows from 265 herds in Ontario; the number of cows sampled within each herd ranged from 1 to 13. The overall prevalence of anti-BIV antibodies among cows was 5.5% with a 95% confidence interval of 4.2% to 7.2%. In contrast, 18.1% of herds had at least one reactor among cows tested, resulting in a herd-prevalence confidence interval of 13.8% to 23.4%. These estimates of prevalence were in the same range as previous reports from the US and Europe. Bovine immunodeficiency-like virus may have a worldwide distribution. Unfortunately, BIV test sensitivity and specificity are difficult to estimate because virus isolation is inefficient. Therefore, the apparent prevalences could not be adjusted for test sensitivity and specificity, to estimate the true prevalence of infection. The serum samples had previously been tested for antibodies to bovine leukemia virus (BLV). There were no significant associations between BIV and BLV test results. Least squares regression was used to investigate potential associations between BIV test results and selected production indices.(ABSTRACT TRUNCATED AT 250 WORDS)