Introduction to network analysis and its implications for animal disease modelling.

Social networks analysis (SNA) has recently been used in veterinary epidemiology to study livestock movements. A network is obtained by considering livestock holdings as nodes in a network and movements among holdings as links among nodes. Social networks analysis enables the study of the network as a whole, exploring all the relationships among pairs of farms. Highly connected livestock holdings in the network can be identified, which can help surveillance and disease prevention activities. Observed livestock movement networks in various countries have shown an important level of contact heterogeneity and clustering (topological, not necessarily geographical or spatial) and understanding the architecture of these networks has provided a better understanding of how infections may spread. The findings of SNA studies of livestock movement should be used to build and parameterise epidemiological models of infection spread in order to improve the reliability of the outputs from these models.

Associations between oil- and gas-well sites, processing facilities, flaring, and beef cattle reproduction and calf mortality in western Canada.

From the fall of 1992 through calving 1996, detailed cow breeding outcome records were maintained actively for seven large cow-calf herds in western Canada. The numbers of mature females in the study for the breeding seasons beginning in 1992, 1993, 1994, and 1995 were 1122, 1177, 1251, and 1236, respectively. Outcomes included pregnancy status, calving interval, and the occurrence of twins, abortions, stillbirths, and neonatal mortality. Information also was collected on other risk factors known to influence beef-herd health and productivity. Detailed maps of active and inactive oil and natural-gas sites, batteries, compressor stations and processing plants were verified. Records of flaring activity at each facility were obtained from the government regulatory agency. Each flaring site then was classified as sour or sweet based on the presence or absence of hydrogen sulfide in the flared gas. A detailed inventory was prepared itemizing the type and number of facilities within 1.6 km (1 mile) of the center of each quarter section used for pasture. The total volume of gas flared within 1.6 km of the center of each pasture was determined for each month of the study. Appropriate risk periods where specified for each outcome and a cumulative exposure calculated for each breeding female (using detailed individual-animal records of cow movements between pastures and herd-management groups). Generalized estimating equations were used to evaluate the association between exposure and outcome and to adjust for potential confounders and clustering of binomial outcomes within herd. Increased risk of non-pregnancy was sometimes associated with exposure to one or more of the following facility types: sour-gas flaring battery facilities, all battery-flaring sites, active gas wells, and larger field facilities. The associations were not, however, consistent among years or even among risk periods for the same year. Facility proximity and flaring were not associated with increased abortion risk. Volume of flared sour gas from battery sites was associated with increased risk of stillbirth. Finally, sour-gas flaring was associated with increased calf-mortality risk for the 1992-1993 calf crop. Several examples of associations between exposure and increased productivity also were found (most of which involved either oil wells or all well sites).

Associations between total sulfation, hydrogen sulfide deposition, and beef-cattle breeding outcomes in western Canada.

Beef cows sometimes are pastured close to oil- and gas-production and processing facilities. We measured the health and productivity of cow-calf herds with differing exposure to sour-natural-gas processing facilities in western Canada. In seven cow-calf herds, the numbers of eligible mature females included in the study for the breeding seasons beginning in 1993, 1994, and 1995 were 1177, 1251, and 1236, respectively. Outcomes included pregnancy status, calving interval, and the occurrence of twins, abortions, stillbirths, and neonatal mortality. Information also was collected on other risk factors known to influence beef-herd health and productivity. Monthly measurements from a network of passive air-monitoring devices were used to estimate exposure. Total sulfation and H2S deposition were used as markers for the complex mixture of compounds found in emissions from sour-gas processing plants and sour flares. Most herds were managed in multiple pasture groups. Cumulative exposure assessments were made from records of individual-animal movements between pastures. Generalized estimating equations were used to evaluate the association between exposure and outcome and to adjust for potential confounders and clustering of binomial outcomes within herd. No consistent associations were found between either total sulfation or H2S deposition and productivity parameters across the cow-calf production cycles. There were, however, five examples of significant associations between increasing cumulative exposure to total sulfation and decreased productivity in the 18 models examined.

Infectious hematopoietic necrosis virus antibody profiles in naturally and experimentally infected Atlantic salmon Salmo salar.

Atlantic salmon Salmo salar naturally and experimentally exposed to infectious hematopoietic necrosis virus (IHNV) in British Columbia, Canada, developed antibodies against the virus. More than 50% of the fish exposed to IHNV remained seropositive for several months after the IHN epizootic had subsided. The virus itself could not be detected in asymptomatic fish once the fish had recovered from IHN. The persistence of IHNV-specific antibodies in a large percentage of Atlantic salmon, from 4 different populations that survived an outbreak of IHN, and the lack of IHNV-specific antibodies in fish with no history of the disease, suggests that serology may be a useful tool for determining previous exposure to the virus. It may be important to determine whether Atlantic salmon have been infected with IHNV because, although the virus is difficult to detect in asymptomatic fish, an incidental finding suggests it may persist in a small number of fish after the outbreak has subsided. Furthermore, the presence of seropositive fish would be an indication that the virus may be enzootic at a farm, and such information would thus aid producers with stocking decisions.

Evidence for a carrier state of infectious hematopoietic necrosis virus in chinook salmon Oncorhynchus tshawytscha.

In British Columbia, Canada, infectious hematopoietic necrosis virus (IHNV) is prevalent in wild sockeye salmon Oncorhynchus nerka and has caused disease in seawater net-pen reared Atlantic salmon Salmo salar. In this study, chinook salmon Oncorhynchus tshawytscha experimentally exposed to an isolate of IHNV found in British Columbia became carriers of the virus. When Atlantic salmon were cohabited with these virus-exposed chinook salmon, IHNV was isolated from the Atlantic salmon. Identification of chinook salmon populations that have been exposed to IHNV may be difficult, as virus isolation was successful only in fish that were concurrently infected with either Renibacterium salmoninarum or Piscirickettisia salmonis. Also, IHNV-specific antibodies were detected in only 2 of the 70 fish experimentally exposed to the virus. Two samples collected from chinook salmon exposed to IHNV while at a salt water net-pen site had a seroprevalence of 19 and 22%; however, the inconsistencies between our laboratory and field data suggest that further research is required before we can rely on serological analysis for identifying potential carrier populations. Because of the difficulty in determining the exposure status of populations of chinook salmon, especially if there is no concurrent disease, it may be prudent not to cohabit Atlantic salmon with chinook salmon on a farm if there is any possibility that the latter have been exposed to the virus.

Observer variation in the histological diagnosis of plasmacytoid leukemia (marine anemia).

The reliability of past histological criteria for the diagnosis of plasmacytoid leukemia was investigated by studying observer variation in the diagnosis of the disease. Participants blindly evaluated a series of histological samples and classified them as positive, negative or questionable cases of the disease. Intra- and interpathologist agreement were used to assess reliability and were determined by calculating the observed agreement in diagnosis and the kappa statistic. There was poor overall agreement for the classification of sample cases. Questionable cases were most frequently misclassified but marked disagreement also existed for positive and negative samples. Access to historical information prior to histological examination influenced agreement. The results show that previously described histological criteria alone were insufficient for establishing the diagnosis of plasmacytoid leukemia.

Efficacy of immunization of feedlot calves with a commercial Haemophilus somnus bacterin.

Two cohorts, consisting of 10,723 calves total, were identified in this prospective follow-up study to investigate whether immunization of auction market beef calves immediately upon arrival at the feedlot with a commercial Haemophilus somnus whole cell killed bacterin would reduce subsequent mortality. In addition to mortality rate, the use of incidence rate of fatal disease is introduced as an effect measure to examine vaccine efficacy in the feedlot. The Haemophilus somnus bacterin had no significant effect on the overall crude mortality rate; however, the bacterin appeared to significantly (p less than 0.05) reduce the incidence rate of fatal disease and the mortality rate during the first two months in the feedlot, when risk of fatal disease onset was highest. Once mortalities likely not associated with hemophilosis (for example, a fractured femoral neck) were removed from the analysis, steer mortality rate, but not heifer mortality rate, was reduced significantly (p less than 0.05) in the vaccinated group. The attributable percent overall for steers was 17.4%; this suggests that 17.4% of fatal respiratory disease in the unvaccinated steers could have been prevented by vaccination with the H. somnus bacterin. Heifer calves demonstrated a significantly (p less than 0.01) higher incidence rate of fatal disease during the first week than did steer calves, indicating that a different pattern of fatal disease existed for the two sexes. Use of a second vaccination two weeks after arrival did little to decrease mortality risk.

Effect of time of year, weather, and the pattern of auction market sales on fatal fibrinous pneumonia (shipping fever) in calves in a large feedlot in Alberta (1985-1988).

A total of 58,885 spring-born calves entering a large commercial feedlot in southwestern Alberta were studied to examine the associations between shipping fever mortality and the pattern of calf sales at the auction markets, time of year, and weather. The observational study followed calves purchased from 42 auction markets in the 4 western provinces between September 1 and December 31 in each of the years from 1985 to 1988. Calf sales at the auction markets consistently peaked during the last week of October and the first week of November. Calves entering the feedlot in November had a risk of fatal shipping fever 2 to 8 times greater than calves entering in September or December. The pattern was the same for all 4 years, with maximum risk occurring 2 to 4 weeks after the peak time for calf sales at the markets. A number of factors could have contributed to this pattern, including changes in transport truck availability, changes in the density of calves at the markets, changes in population dynamics at the feedlot that affected feedlot crew efficiency, and weather. The finding that the risk of fatal shipping fever appears to increase significantly as the feedlot fills with calves in the fall deserves the attention of feedlot owners, so they can design their treatment strategies appropriately, and of researchers, who may gain useful knowledge about the natural history of the disease by investigating why this change in risk occurs.

Epidemiological study of enzootic pneumonia in dairy calves in Saskatchewan.

A field study involving 325 calves from 17 dairy herds in Saskatchewan was conducted to determine the risk of enzootic pneumonia and to assess its association with a number of factors. Two different case definitions of pneumonia were used in the analyses: the first was based on producers' treatment risk (CASE1) and the second was based on semimonthly clinical examinations of calves by the research veterinarian (CASE2). The risk of pneumonia based on CASE1 was 39% and on CASE2 was 29%. The measure of agreement between CASE1 and CASE2 at the calf level of analysis was poor (kappa = 0.24, SE = 0.02) and at the herd level of analysis was moderate (kappa = 0.40, SE = 0.12). The mortality risk from pneumonia was 1.8% and a variety of infectious organisms were isolated from pneumonic lungs. Twenty-seven percent of the calves had inadequate (total IgG < or = 800 mg/dL) levels of passively acquired antibodies as measured by radial immunodiffusion. The proportion of seropositive titers in calves within the first two weeks of age was 94% to parainfluenza 3 virus (PI3V) and bovine respiratory syncytial virus (BRSV), 73% to Pasteurella haemolytica (Ph), 68% to bovine viral diarrhea virus (BVDV), 67% to infectious bovine rhinotracheitis virus (IBRV), 46% to Mycoplasma dispar (Md), 44% to Haemophilus somnus (Hs), and 21% to Mycoplasma bovis (Mb). At the calf level of analysis and after adjusting for clustering, there was a negative association (p = 0.10) between the diagnosis of pneumonia based on CASE2 and total IgG levels and Ph titers (rPh).(ABSTRACT TRUNCATED AT 250 WORDS)

The predictive value of pelvimetry in beef cattle.

To elucidate reasons for failure of pelvimetry to predict dystocia, we collected data from 1146 heifers and 210 cows in five beef cow herds in Saskatchewan. We assessed the reliability of pelvic area measurements, the generalizability of findings, various modifications of the technique, and the statistical association between pelvic area measurements and dystocia. The repeatability (kappa) of pelvic area measurements between and within veterinarians for the Rice and Krautmann pelvimeters were low to moderate, indicating pelvic area measurements were imprecise. The positive predictive values and sensitivities of pelvic area measurements were consistently poor across herds, years of study, breeds of heifers, times of measurement, various pelvic area cut-off points, and sires. Various modifications of the technique, including pelvic area/calf birth weight ratios, pelvic area/heifer weight ratios, and Ko's calving prediction equation were also poor on-farm tests for predicting dystocia. Although the mean pelvic area in heifers with dystocia was smaller than those without dystocia, there was a large overlap in the distribution of their measurements. Far too many heifers with a small pelvic area had no dystocia (false positives) and far too many heifers with a large pelvic area had dystocia (false negatives) for pelvimetry to be useful. We conclude there is little evidence to justify the continued use of pelvimetry as an on-farm test to reduce dystocia in beef cattle.

The prevalence of bovine viral diarrhea virus infection in a population of feedlot calves in western Canada.

The prevalence of bovine viral diarrhea virus (BVDV) infection was examined in a population of 5129 recently weaned steer calves entering a large feedlot in central Saskatchewan from September to December 1991. Serum samples were collected within 24 h of arrival at the feedlot from every fifth calf processed and again 96 d postarrival. A microtiter virus isolation test was used to determine the prevalence of calves viremic with BVDV on entry to the feedlot. An enzyme-linked immunosorbent assay (ELISA) which detects antibody against glycoprotein 53 of the BVDV was used on paired sera to determine the seroconversion risk during the first 96 d in the feedlot. A virus neutralization (VN) test for BVDV was conducted on a sub-sample of paired sera to measure agreement in determination of seroconversion risk with the ELISA. A polymerase chain reaction (PCR) test which detects BVDV was used to determine if cattle were acutely viremic when treated for disease. The estimated prevalence of persistently infected calves in this population was < 0.1%. The seroconversion risk for BVDV was 27% (236/864) according to the ELISA and it varied from 0 to 63% among the 20 pens sampled. According to the VN test, the seroconversion risk for BVDV was 40% (132/327) and it varied from 0 to 100% among the 11 pens tested. The agreement between the ELISA and VN tests in seroconversion risk to BVDV was very poor (kappa = 0.15 +/- 0.039 SE). The prevalence of acute viremia in calves treated at the feedlot hospital was low at 4% (6/149).(ABSTRACT TRUNCATED AT 250 WORDS)

Comparison of two sampling tools for diagnosis of Tritrichomonas foetus in bulls and clinical interpretation of culture results.

OBJECTIVE: To compare sensitivity for diagnosing Tritrichomonas foetus infection in bulls using 2 sampling tools and to calculate negative predictive values for infection. DESIGN: Randomized clinical trial. ANIMALS: 30 Bos taurus bulls naturally or experimentally infected with T foetus. PROCEDURE: Preputial scrapings were obtained once/wk for 6 weeks using an artificial insemination pipette and a metal brush; which tool was used first for each bull was randomly determined. Samples were collected first from the left side of the prepuce and then from the right side and placed in commercially available transport media chi 2 Values and confidence limits were adjusted for effect of clustering of results by bull. RESULTS: Significant differences in sensitivity of results were not found between samples collected using the brush or pipette. Using the pipette, sensitivity was estimated to be 91.6% (95% confidence interval, 84.3 to 95.7%); negative predictive values ranged from 41 to 99% for prevalence of infection of 90 to 5%, respectively. Sensitivity was 88.8% for first sample obtained and 96.1% for second sample obtained. CONCLUSIONS AND CLINICAL RELEVANCE: Collection of preputial scrapings with an artificial insemination pipette or a metal brush and use of a commercially available culture system can provide a sensitive diagnostic test for T foetus infection in bulls. Calculated negative predictive values indicated that 1 or 2 tests would suffice in most clinical situations. For bulls from herds in which T foetus is endemic, 2 to 4 tests/bull may be required to ensure that each bull is not infected.

Determination of the association between Neospora caninum infection and reproductive performance in beef herds.

OBJECTIVE: To determine seroprevalence of Neospora caninum infection in beef herds and the association between serologic status and rate of abortion, stillbirth, calf mortality, and reproductive failure. DESIGN: Longitudinal study. ANIMALS: 419 cows from 8 beef herds in central Alberta. PROCEDURE: 1,391 serum samples from a serum bank were analyzed, using ELISA, and results were compared, using logistic regression, with productivity data on individual cows obtained from a database established during a 4-year period. RESULTS: 30% of cows were seropositive at some point during the 4-year period. Risks of abortion (odds ratio [OR], 5.7) and stillbirth (OR, 2.8) in seropositive cows were significantly greater than in seronegative cows. Risks of being culled for any reason (OR, 1.9) or for reproductive failure (OR, 2.5) in seropositive cows were also significantly greater than in seronegative cows. Changes in titer with time in individual cows and a lack of association between serologic status of dam versus daughter suggest that postnatal transmission was possible in these herds. However, horizontal transmission did not appear to play a substantial role in abortions that occurred in these herds. CLINICAL IMPLICATIONS: Neosporosis should be investigated as a potential source of economic loss to the beef industry.

Risk factors associated with parvovirus enteritis in dogs: 283 cases (1982-1991).

OBJECTIVE: To determine breed, sex, and seasonal predisposition for development of canine parvovirus (CPV) enteritis in dogs. DESIGN: Retrospective case-control study. SAMPLE POPULATION: Medical records from 283 dogs with confirmed CPV enteritis and from 834 age-matched control dogs that were healthy or had been admitted with nonenteric illness. PROCEDURE: Effects of season, breed, sex, and neutering on the risk of developing CPV enteritis were examined by calculation of unadjusted odds ratios and performance of multivariate analysis. Stratified and contingency table analyses were performed to identify interactions and confounding among variables. RESULTS: Rottweilers, American Pit Bull Terriers, Doberman Pinschers, and German Shepherd Dogs were at increased risk and Toy Poodles and Cocker Spaniels were at decreased risk for developing CPV enteritis, compared with that for mixed-breed dogs. For dogs more than 6 months old, sexually intact males were twice as likely as intact females to develop CPV enteritis. Dogs were 3 times more likely to be admitted with CPV enteritis in July, August, and September, compared with the rest of the year. Dogs were 12.7 times more likely to be admitted with CPV enteritis if they had not been currently vaccinated. CLINICAL IMPLICATIONS: Lack of vaccination is a significant risk factor for development of CPV enteritis. Seasonal, sex, and breed predispositions for the development of CPV enteritis also exist.

Evaluation of the impact of a natural gas leak from a pipeline on productivity of beef cattle.

OBJECTIVE: To determine the association between a leak of sour natural gas (more than 30% hydrogen sulfide) from a pipeline in a river valley and the health of beef cattle in the intensively ranched surrounding area. DESIGN: Prospective cohort study. SAMPLE POPULATION: 13 herds of cattle within 4 km (2.5 miles) of the leak and 10 herds outside the 4-km zone. PROCEDURE: Distance of herds from the leak site was determined, using geographic information system technology. Information about speed and direction of winds was obtained from a local meteorologic station and an ambient air-quality monitoring trailer. Health and productivity data for surrounding beef herds, as well as exposure information, were collected and analyzed. RESULTS: An association was not found between total herd calf mortality and herd distance from the leak, wind-aided exposure, location in the river valley, signs of irritation consistent with exposure to the gas, or reports of odors of gas at the time of the leak. Management changes reported in response to the gas leak were identified as risk factors for total herd calf mortality. Other herd-level risk factors associated with increased calf mortality ratio included a median calving date in February and percentage of twin births for a herd. CLINICAL IMPLICATIONS: In this example, we did not detect an association between productivity of cattle and exposure to sour natural gas. Several methods can be used for ranking potential exposure after discovery of a leak.

Effect of transportation on fatal fibrinous pneumonia and shrinkage in calves arriving at a large feedlot.

A retrospective, epidemiologic field study was performed to determine whether calves that were transported greater distances from the auction market to the feedlot were at significantly higher risk of developing fatal fibrinous pneumonia (shipping fever). The study involved all 45,243 springborn steer calves that were purchased from auction markets and moved into a large commercial feedlot between September 1 and December 31 over a 4-year period (1985-1988). For all 4 years of the study, the distance calves were transported from the market to the feedlot and their shrinkage or subsequent death loss from fibrinous pneumonia were not correlated. The risk of fatal fibrinous pneumonia for calves arriving from nearby markets was just as high as that for calves transported much greater distances. The results suggested that calves can be purchased from more distant markets without having to discount their price for higher expected death losses. Differences between short and long hauls explained little, if any, of the variation among truckloads of calves in the risk of fatal fibrinous pneumonia.

Effect of pretransit mixing on fatal fibrinous pneumonia in calves.

A retrospective, epidemiologic study was performed to quantify the mixing of calves from various sources at auction markets, and to determine whether mixing at the markets and the risk of fatal fibrinous pneumonia (FFP) at the feedlot were associated. In this study, 32,646 spring-born steer calves that entered a single large feedlot during the fall seasons between 1985 and 1988 were traced back to their originating auction market, and sales tickets were used to measure the number of farm sources that contributed to each truckload of calves. Individual cow/calf producers contributed a median of only 2 calves/truckload arriving at the feedlot in this study. An average truckload of 60 steers comprised calves from as many as 20 to 30 farms. The degree of mixing varied little over time and could not, therefore, be used to explain the large variations in FFP risk during different months and different years of the study. However, variation in the degree of mixing of calves from various sources at the markets evidently was responsible for differences in FFP risk among truckloads assembled by different buyers. When truckloads were grouped by buyer, FFP risk and mean number of calves per source were negatively correlated in 1986 (r = -0.67, P = 0.099), and in 1987 (r = -0.90, P = 0.002). These variables also were negatively correlated in 1988 (r = -0.56), although the correlation was not significant. The positive linear relationship between mixing of calves in truckloads supplied by different buyers and subsequent FFP risk suggested that veterinarians and feedlot owners should more aggressively observe and treat calves from truckloads that were highly mixed.(ABSTRACT TRUNCATED AT 250 WORDS)

Survey of veterinary technical and professional skills in students and recent graduates of a veterinary college.

OBJECTIVES: To determine perceptions of veterinary technical and professional skills among veterinary students and recent graduates. DESIGN: Cross-sectional study. SAMPLE POPULATION: 281 students and 142 recent graduates from the Ontario Veterinary College. PROCEDURE: A survey was designed and administered to first- through fourth-year students and veterinarians who had graduated either 1 or 6 years before survey administration. RESULTS: Overall response rate was 70%. Learning about technical and professional skills was highly valued. Most participants felt they had not received instruction about professional skills, but those who had felt more competent about them. Perceptions of competence increased slightly with increased comfort discussing emotional veterinary issues with instructors. Neither gender nor increased age was related to increased feelings of competence. Almost all fourth-year students felt competent and comfortable about examining an animal with the client present, assessing suffering, diagnosing parvovirus infection, performing surgery, and working as group members. However, many did not feel competent or comfortable about delivering bad news, setting time limits yet providing quality service, helping clients with limited funds make treatment decisions, dealing with demanding people, and euthanasia. Feelings of competence and comfort were closely related but were not identical. CONCLUSIONS AND CLINICAL RELEVANCE: In the interests of best preparing entry-level veterinarians, technical and professional skills need to be emphasized in a learning environment where students feel comfortable discussing emotional veterinary issues. A professional skills curriculum addressing underlying self-awareness, communication, and interpersonal issues, as well as procedural matters, would likely increase the proportion of fourth-year students who feel competent and comfortable about professional skills by the end of their undergraduate training.

Developmental Anomalies in Farm Animals: II. Defining Etiology.

Anomalous development is a function of the genetic make-up of the fetus and the environment in which it develops. The cause of anomalous development may occasionally be obvious but more often is obscure, because of its multifactorial nature. Therefore a methodological approach is often required to define the etiology. In this paper we describe a methodological approach that first defines the incident, the genotype of the individuals involved, and the environment in which the fetus has developed. This procedure can be undertaken in a stepwise fashion and need not be followed through to the end if the etiology becomes obvious. This approach highlights the necessity to characterize the defects, to use ancillary tests, and to apply basic epidemiological methods.

The effects of changing demographics on the distribution of marine anemia in farmed salmon in British Columbia.

The changing geographic distribution of marine anemia (plasmacytoid leukemia) was compared with the evolving demographics of the chinook farming industry in British Columbia to explore the hypothesis that the disease had spread throughout the province between 1987 and 1992. Through retrospective and prospective methods, it was shown that the apparent spread of the disease was likely an artifact, resulting from changes in the distribution of fish farms throughout the province and corresponding changes in the intensity of regional disease surveillance. When viewed over a 5-year period, there were no statistically significant differences in the prevalence of the disease amongst fish sampled from each of the major salmon farming regions of British Columbia. By increasing the intensity of surveillance for the disease in apparently negative regions or chinook farms, one could routinely find cases of the disease. The results suggest that marine anemia is an endemic problem for farmed chinook salmon in British Columbia and is not a spreading epidemic.