Who let the dogs out? Exploring the spatial ecology of free-roaming domestic dogs in western Kenya.

The spatial ecology of free-roaming dogs determines their role in the transmission of zoonoses. This study describes the geographic range of and identifies sites frequently visited by free-roaming domestic dogs in western Kenya. Eight sites in Busia county, western Kenya, were selected. At each site, ten dog-keeping households were recruited, a questionnaire was administered, and a GPS logger was fixed around the neck of one dog in each household. Loggers were programmed to capture the dog's position every minute, for five consecutive days. Individual summaries of GPS recordings were produced, and the daily distance traveled was calculated. 50% and 95% utilization distribution isopleths were produced, and the area within these isopleths was extracted to estimate the size of the core and extended Home Ranges (HRs), respectively. Linear regression analyses were performed to identify factors associated with the movement parameters. The centroid points of the 10, 50, and 90% isopleths were reproduced, and the corresponding sites identified on the ground. Seventy-three dogs were included in the final analyses. The median daily distance traveled was 13.5km, while the median core and extended HRs were 0.4 and 9.3 ha, respectively. Older dogs had a larger extended HR and traveled more daily, while the effect of sex on dog movement depended on their neutering status. Dogs spent most of their time at their household; other frequently visited sites included other household compounds, fields, and rubbish dumps. One of the centroids corresponded to a field located across the international Kenya-Uganda border, emphasizing the fluidity across the border in this ecosystem. Multiple dogs visited the same location, highlighting the heterogeneous contact networks between dogs, and between dogs and people. The field data presented are of value both in understanding domestic dog ecology and resource utilization, and in contextualizing infectious and parasitic disease transmission models.

Multiple imputation in veterinary epidemiological studies: a case study and simulation.

The problem of missing data occurs frequently in veterinary epidemiological studies. Most studies use a complete case (CC) analysis which excludes all observations for which any relevant variable have missing values. Alternative approaches (most notably multiple imputation (MI)) which avoid the exclusion of observations with missing values are now widely available but have been used very little in veterinary epidemiology. This paper uses a case study based on research into dairy producers' attitudes toward mastitis control procedures, combined with two simulation studies to evaluate the use of MI and compare results with a CC analysis. MI analysis of the original data produced results which had relatively minor differences from the CC analysis. However, most of the missing data in the original data set were in the dependent variable and a subsequent simulation study based on the observed missing data pattern and 1000 simulations showed that an MI analysis would not be expected to offer any advantages over a CC analysis in this situation. This was true regardless of the missing data mechanism (MCAR - missing completely at random, MAR - missing at random, or NMAR - not missing at random) underlying the missing values. Surprisingly, recent textbooks dealing with MI make little reference to this limitation of MI for dealing with missing values in the dependent variable. An additional simulation study (1000 runs for each of the three missing data mechanisms) compared MI and CC analyses for data in which varying levels (n=7) of missing data were created in predictor variables. This study showed that MI analyses generally produced results that were less biased on average, were more precise (smaller SEs), were more consistent (less variability between simulation runs) and consequently were more likely to produce estimates that were close to the "truth" (results obtained from a data set with no missing values). While the benefit of MI varied with the mechanism used to generate the missing data, MI always performed as well as, or better than, CC analysis.

Dealing with deficient and missing data.

Disease control decisions require two types of data: data describing the disease frequency (incidence and prevalence) along with characteristics of the population and environment in which the disease occurs (hereafter called "descriptive data"); and, data for analytical studies (hereafter called "analytical data") documenting the effects of risk factors for the disease. Both may be either deficient or missing. Descriptive data may be completely missing if the disease is a new and unknown entity with no diagnostic procedures or if there has been no surveillance activity in the population of interest. Methods for dealing with this complete absence of data are limited, but the possible use of surrogate measures of disease will be discussed. More often, data are deficient because of limitations in diagnostic capabilities (imperfect sensitivity and specificity). Developments in methods for dealing with this form of information bias make this a more tractable problem. Deficiencies in analytical data leading to biased estimates of effects of risk factors are a common problem, and one which is increasingly being recognized, but options for correction of known or suspected biases are still limited. Data about risk factors may be completely missing if studies of risk factors have not been carried out. Alternatively, data for evaluation of risk factors may be available but have "item missingness" where some (or many) observations have some pieces of information missing. There has been tremendous development in the methods to deal with this problem of "item missingness" over the past decade, with multiple imputation being the most prominent method. The use of multiple imputation to deal with the problem of item missing data will be compared to the use of complete-case analysis, and limitations to the applicability of imputation will be presented.

Sensitivity and specificity of repeated test results from a commercial milk enzyme-linked immunosorbent assay for detection of Mycobacterium avium subspecies paratuberculosis in dairy cattle.

OBJECTIVE: To investigate the sensitivity and specificity of results of initial and repeated milk ELISAs (at 6- or 12-month intervals) to detect cows that were shedding Mycobacterium avium subsp paratuberculosis (ie, were infectious) and to evaluate factors influencing the probability that the results of a repeated milk ELISA would be positive for an infectious cow if the results of the initial milk ELISA were negative. DESIGN: Prospective cohort study. ANIMALS: 3,145 dairy cows from 32 herds. PROCEDURES: Herds from the 3 Maritime provinces in Canada (Prince Edward Island, New Brunswick, and Nova Scotia), participating in a Dairy Herd Improvement program, and that had undergone a prior Mycobacterium avium subsp paratuberculosis awareness project were selected for the study. Sample collection occurred between April 2009 and March 2011 with milk and fecal samples collected from all lactating cows in study herds every 6 months. Herds completing < 3 herd visits with collection of individual cow fecal or milk samples, within this sampling timeframe, were excluded from analyses. Fecal samples were cultured in liquid medium and a cow was defined as infectious if ≥ 1 sample was culture positive (reference test). A milk ELISA (index test) was completed with a commercial kit, following manufacturer's instructions. RESULTS: For a 6-month test interval, sensitivities of the milk ELISA to detect infectious cows were 22.0% and 32.6% for initial and combined initial and repeated tests (parallel interpretation), respectively. Specificity of the initial ELISA was 99.6% and was 99.2% for combined tests. For a 12-month test interval, sensitivities of the milk ELISA to detect infectious cows were 25.6% and 45.3% for initial and combined initial and repeated tests (parallel interpretation), respectively. Specificity of the initial ELISA was 99.6% and was 98.9% for combined tests. In infectious cows, magnitude of the initial negative ELISA result was a positive predictor for a positive repeated ELISA result. CONCLUSIONS AND CLINICAL RELEVANCE: Results of a repeated milk ELISA improved detection of Mycobacterium avium subsp paratuberculosis infectious cows, with minimal loss of specificity. A 12-month test interval provided a greater increase in sensitivity, relative to an initial test, than did a 6-month interval. Infectious cows with an initial negative milk ELISA result close to the cutoff for a positive test were more likely to have positive results on a repeated ELISA. Repeated testing improved detection of infectious cows and reduced risk of misclassification compared with a single ELISA result.

Bias--is it a problem, and what should we do?

Observational studies are prone to two types of errors: random and systematic. Random error arises as a result of variation between samples that might be drawn in a study and can be reduced by increasing the sample size. Systematic error arises from problems with the study design or the methods used to obtain the study data and is not influenced by sample size. Over the last 20 years, veterinary epidemiologists have made great progress in dealing more effectively with random error (particularly through the use of multilevel models) but paid relatively little attention to systematic error. Systematic errors can arise from unmeasured confounders, selection bias and information bias. Unmeasured confounders include both factors which are known to be confounders but which were not measured in a study and factors which are not known to be confounders. Confounders can bias results toward or away from the null. The impact of selection bias can also be difficult to predict and can be negligible or large. Although the direction of information bias is generally toward the null, this cannot be guaranteed and its impact might be very large. Methods of dealing with systematic errors include: qualitative assessment, quantitative bias analysis and incorporation of bias parameters into the statistical analyses.

Evaluation of environmental fecal culture for Mycobacterium avium subspecies paratuberculosis detection in dairy herds and association with apparent within-herd prevalence.

This study evaluated test characteristics of environmental culture (EC) for the detection of Mycobacterium avium subspecies paratuberculosis (MAP) in 32 herds over a 2-year period. Individual fecal samples were collected every 6 mo and environmental samples every 3 mo. Individual fecal culture was performed on samples from positive pools. Samples were cultured in broth, with confirmatory polymerase chain reaction performed on positive fecal samples. Repeated measures were accounted for using GEE logistic models. Relative to a MAP herd-status based on all pooled fecal culture results collected during the study, sensitivity of a set of 6 EC-samples collected from prescribed locations within the herd environment (EC-6) was 71% [95% confidence interval (CI): 49% to 86%] and specificity was 99% (95% CI: 95% to 100%). Sensitivity of EC increased as apparent within-herd fecal culture prevalence (aWHP) increased. The estimated aWHP increased as the proportion of positive EC-samples within an EC-6 set increased. Environmental culture is an acceptable tool for herd diagnosis of MAP in low-prevalence herds.

Évaluation de la culture fécale environnementale pour la détection deMycobacterium aviumsous-espèceparatuberculosisdans les troupeaux laitiers et l'association avec la prévalence apparente dans les troupeaux. Cette étude a évalué les caractéristiques des tests de cultures environnementales (CE) pour la détection de Mycobacterium avium sous-espèce paratuberculosis (MAP) dans 32 troupeaux pendant une période de deux ans. Des échantillons fécaux individuels ont été prélevés tous les 6 mois et des échantillons environnementaux tous les 3 mois. La culture fécale individuelle a été réalisée sur des échantillons provenant des échantillons regroupés positifs. Les échantillons ont été cultivés dans du bouillon et l'amplification en chaîne par la polymérase de confirmation a été réalisée sur des échantillons fécaux positifs. Des mesures répétées ont été enregistrées à l'aide de modèles logistiques d'équations généralisées d'estimation. En rapport avec un statut de troupeau pour MAP fondé sur tous les résultats des cultures fécales regroupées prélevées durant l'étude, la sensibilité d'un groupe de 6 échantillons-CE prélevés dans les lieux prescrits au sein de l'environnement du troupeau (CE-6) était de 71 % [intervalle de confiance (IC) de 95 %: de 49 % à 86 %] et la spécificité était de 99 % (IC de 95 %: de 95 % à 100 %). La sensibilité de CE a augmenté au fur et à mesure que la prévalence apparente de la culture fécale dans le troupeau (PaCF) montait. La PaCF estimée augmentait tandis que la proportion d'échantillons CE positifs dans le groupe CE-6 montait. La culture environnementale est un outil acceptable pour le diagnostic de MAP chez le troupeau pour les troupeaux avec une faible prévalence.(Traduit par Isabelle Vallières).

Differences in behavioral characteristics between dogs obtained as puppies from pet stores and those obtained from noncommercial breeders.

OBJECTIVE: To compare the owner-reported prevalence of behavioral characteristics in dogs obtained as puppies from pet stores with that of dogs obtained as puppies from noncommercial breeders. DESIGN: Cross-sectional study. Animals-Dogs obtained as puppies from pet stores (n = 413) and breeder-obtained dogs (5,657). PROCEDURES: Behavioral evaluations were obtained from a large convenience sample of current dog owners with the online version of the Canine Behavioral Assessment and Research Questionnaire, which uses ordinal scales to rate either the intensity or frequency of the dogs' behavior. Hierarchic linear and logistic regression models were used to analyze the effects of source of acquisition on behavioral outcomes when various confounding and intervening variables were controlled for. RESULTS: Pet store-derived dogs received significantly less favorable scores than did breeder-obtained dogs on 12 of 14 of the behavioral variables measured; pet store dogs did not score more favorably than breeder dogs in any behavioral category. Compared with dogs obtained as puppies from noncommercial breeders, dogs obtained as puppies from pet stores had significantly greater aggression toward human family members, unfamiliar people, and other dogs; greater fear of other dogs and nonsocial stimuli; and greater separation-related problems and house soiling. CONCLUSIONS AND CLINICAL RELEVANCE: Obtaining dogs from pet stores versus noncommercial breeders represented a significant risk factor for the development of a wide range of undesirable behavioral characteristics. Until the causes of the unfavorable differences detected in this group of dogs can be specifically identified and remedied, the authors cannot recommend that puppies be obtained from pet stores.

Effectiveness of emamectin benzoate for treatment of Lepeophtheirus salmonis on farmed Atlantic salmon Salmo salar in the Bay of Fundy, Canada.

Emamectin benzoate (an avermectin chemotherapeutant administered to fish as an in-feed treatment) has been used to treat infestations of sea lice Lepeophtheirus salmonis on farmed Atlantic salmon Salmo salar in the Bay of Fundy, New Brunswick, Canada, since 1999. This retrospective study examined the effectiveness of 114 emamectin benzoate treatment episodes from 2004 to 2008 across 54 farms. Study objectives were to establish whether changes in the effectiveness of emamectin benzoate were present for this period, examine factors associated with treatment outcome, and determine variables that influenced differences in L. salmonis abundance after treatment. The analysis was carried out in 2 parts: first, trends in treatment effectiveness and L. salmonis abundance were explored, and second, statistical modelling (linear and logistic regression) was used to examine the effects of multiple variables on post-treatment abundance and treatment outcome. Post-treatment sea lice abundance increased in the later years examined. Mean abundance differed between locations in the Bay of Fundy, and higher numbers were found at farms closer to the mainland and lower levels were found in the areas around Grand Manan Island. Treatment effectiveness varied by geographical region and decreased over time. There was an increased risk for unsuccessful treatments in 2008, and treatments applied during autumn months were more likely to be ineffective than those applied during summer months.

Bulk tank milk selenium and its association with milk production parameters in Canadian dairy herds.

An observational study was conducted in 63 Canadian dairy farms to evaluate the association of bulk tank milk selenium (Se) concentration (BTSe) with average daily milk yield, milk fat and milk protein, bulk tank somatic cell count (BTSCC), and the probability of being a Staphylococcus aureus-positive herd. Bulk tank milk samples collected between March 2007 and February 2008 were evaluated for BTSCC, S. aureus culture status, and BTSe. Mean BTSe was 0.51 ± 0.15 µmol/L; no herds were classified as deficient or marginal based on BTSe. Bulk tank milk Se was unconditionally associated with milk production; however, adjusting by region indicated a confounding effect of this variable. There was no relationship between BTSe and BTSCC. Higher values of BTSe were associated with lower risk of being a S. aureus-positive herd, possibly as a result of a more robust udder immune response, or as a result of providing Se at a higher rate as one component of an udder health program that reduces risk of being S. aureus positive.

Selection of a cutoff value for real-time polymerase chain reaction results to fit a diagnostic purpose: analytical and epidemiologic approaches.

Diagnostic laboratories frequently select a subjective cutoff value for real-time amplification assays, above which a threshold cycle (Ct) value is deemed false. Commonly, higher Ct values are interpreted as amplification or fluorescence artifacts, or cross contaminations. Although the implementation of Ct cutoff might be reasonable, its justification and selection should be based on evidence. The current article reviewed evidence-based strategies to select Ct cutoffs grouped in analytical and epidemiologic approaches. Analytical strategies use criteria gathered during the assay development and include fluorescence threshold, reaction end-cycle, limit of detection, and artifact investigation. Variability in amplification efficacy across test runs may induce some instability in an intended Ct cutoff and requires some standardization or normalization procedures. Epidemiologic strategies use criteria based on either the probability or the cost of a false test result associated with a specified cutoff. Cutoffs, depending on the intended purpose of the test, can be selected graphically to minimize the probability of either false-positive or false-negative results by using two-graph receiver operating characteristics curves. The assay's diagnostic sensitivity and specificity may vary with the tested population, thus, the estimated two-graph receiver operating characteristics curve is population dependent and should be established for the targeted population. Although the selection of a cutoff based on misclassification cost depends on infection prevalence, the selection based on predictive values does not. To optimize the test average diagnostic performance, the Ct cutoff should be selected when diagnostic odds ratio is maximal. Epidemiologic approaches were illustrated by selecting Ct cutoffs for a real-time assay for Infectious salmon anemia virus.

Development and validation of a bilingual questionnaire for measuring udder health related management practices on dairy farms.

Questionnaires are frequently used instruments to collect data in epidemiological studies. In countries where more than one language is spoken, the development of a questionnaire in more than one language is needed. The objective of this study was to develop and test the repeatability and validity of English and French versions of a personal interview-format questionnaire designed to capture udder health related management practices used on dairy farms. A standardized protocol was used to develop and translate the research instrument. Equivalence of the English and French questionnaires was assessed using a cross-over study design with 24 bilingual dairy producers completing both versions on three different occasions in a randomly assigned sequence. Repeatability of the questionnaire was evaluated using the test-retest method with the same questions being asked on two different occasions to 88 dairy producers participating in the National Cohort of Dairy Farms of the Canadian Bovine Mastitis Research Network. Validity of the questions related to milking procedures and general housing was assessed using on-farm observations as a gold standard. Measures of agreement were calculated using kappa, quadratic-weighted Kappa and concordance correlation coefficients for categorical, ordinal and continuous variables, respectively. Sensitivity and specificity estimates were computed for the validity analysis. The overall equivalence of the English and French versions of the questionnaire was adequate; agreement measures when administered twice in the same language were not significantly higher than when administered in each language. Similarly, questionnaire overall repeatability was good. When accounting for prevalence bias, Kappa and CCC estimates ranged from 0.40 to 0.92 for 27 of the 29 items evaluated in the questionnaire, with 18 items yielding agreement estimates greater than 0.60. Finally, milking procedures and general housing questions validity was excellent with mean sensitivity and specificity of 86% and 92%, respectively. Although the overall evaluation of the instrument was satisfactory, specific doubtful items were identified. This illustrates the need to address questionnaire reliability as even rigorously designed and pre-tested questions can have poor repeatability or validity. Our results indicate that the developed English and French questionnaires can be used simultaneously to accurately measure the udder health related management practices used on Canadian dairy farms. This questionnaire is adaptable for use in other developed dairy industry populations. The questionnaire is freely available online at www.mastitisnetwork.org under the "Publications/others" section.

White-line disease and haemorrhages in hooves of Finnish dairy cattle.

The Finnish Healthy Hooves Project was set up to determine the frequency of, and risk factors for, various hoof lesions in Finnish dairy herds. Data were collected in 2003 and 2004. A large data set of >74,000 cow-level observations recorded by hoof-trimmers were merged with production data from the Finnish Agricultural Data Processing Centre Ltd. Ultimately, data from a single lactation from each of 16,727 cows in 703 herds were used for analyses of white-line disease (WLD) and haemorrhages-lesions. Three-level hierarchical logistic models with hoof-trimmer and farms (within hoof-trimmer) as random-effects were fit to datasets of tie-stall (TS) and loose-housing (LH) herds separately. Cows examined once had a WLD risk of 4.8% in TS herds and 17% in LH herds. As the number of examinations increased, the odds of a WLD diagnosis increased substantially; (2 and 3+ examinations had odds ratios (ORs) of 2.57 and 3.40 in TS herds and 2.32 and 4.67 in LH herds, respectively, compared to 1 examination). Parity had an interaction with breed in all models in TS and in the LH haemorrhages model, but not in WLD LH models. In TS herds, older Holstein cows had a very high risk of WLD compared to young Ayrshire cows (OR=7.92). Also in LH herds both breed and parity had association with WLD (Holstein cows were 1.57 times more likely than Ayrshire cows and cows in parity >or=4 were more likely (OR=2.89) than parity 1 to get WLD). In TS herds, other hoof lesions - such as haemorrhages (OR=1.65), heel-horn erosions (OR=1.77), and corkscrew claw (OR=1.82) - increased the risk of WLD. In LH herds, corkscrew claw (OR=1.59) and heel-horn erosion (OR=0.72) had a significant association with WLD. In TS herds, use of mats (compared to hard flooring) significantly reduced the risk of WLD and haemorrhages (OR=0.57 and OR=0.80, respectively). No significant associations of the 305-day milk production on the risk of either disease were observed in either TS or LH herds (305-day milk yield from each cows' previous lactation were used for calculations, with herd average yield and the individual cow's deviation from the herd average being used as predictors).

Use of register data to assess the association between use of antimicrobials and outbreak of Postweaning Multisystemic Wasting Syndrome (PMWS) in Danish pig herds.

In 2001, the first case of Postweaning Multisystemic Wasting Syndrome (PMWS) was reported in the Danish pig population. During subsequent years, the number of affected farms increased exponentially. The aim of this study was to determine how this increase influenced the use of antimicrobials between 2002 and 2004. We used national register data of herd characteristics, antimicrobial usage and disease occurrence. The analysis included data on antimicrobial usage in 3371 pig herds with weaners and 7434 pig herds with finishers, which accounted for 56 and 82% of the national amount of antimicrobials prescribed to weaners (prescribed by 347 practitioners) and finishers (prescribed by 522 practitioners), respectively. The estimation of the effect of PMWS was done by comparing the amount of antimicrobials (measured as Animal Defined Daily Doses (ADDkg) used per pig-day at risk each month in each herd) used in herds before and after an outbreak of PMWS, and by comparing the amount of antimicrobials used in herds experiencing PMWS with the amount of antimicrobials used in herds not experiencing PMWS. The effects were estimated in a three-level (veterinarian/herd/study-month) linear mixed regression model with an autoregressive correlation of order 1 (AR1). We found that after a herd had experienced an outbreak of PMWS, the antimicrobial usage in weaners was increased for a year. During the first 3 months post outbreak the usage increased by 22%, followed by an increase of 7% during the next 4th to 12th month when compared to the pre-outbreak usage. There was a significant variation between herds in this effect. Additionally, in herds experiencing an outbreak of PMWS, the usage of antimicrobials before the outbreak was 37 and 19% higher in herds with weaners and finishers, respectively, compared to herds not experiencing PMWS. Generalisation of the results to the entire Danish pig population indicated that the increase of PMWS infected herds from almost zero to about 20% during a 4-year period resulted in a national increase of 4-5% in antimicrobials usage in weaners. The effect of PMWS on usage of antimicrobials in finishers was unclear.

Sole ulcers in Finnish dairy cattle.

The Finnish Healthy Hooves project was set up to determine the frequency of, and risk factors for various hoof lesions in Finnish dairy herds. Data were collected in the years 2003 and 2004. A large dataset of over 74,000 cow-level observations recorded by hoof trimmers was merged with production data from the Finnish Agricultural Data Processing Centre Ltd. Ultimately, data from a single lactation from each of 16,792 cows in 703 herds were used for the analyses in this paper. Three-level hierarchical logistic models with hoof trimmer and farms (within hoof trimmer) as random effects were fit to data sets of tie stall (TS) and loose housing (LH) herds separately. The outcome of interest was the presence or absence of a sole ulcer in one or more legs of a cow during the lactation of interest. Cows examined once had a risk of sole ulcer 5.23% in tie stall herds and 7.58% in LH herds. As the number of examinations increased the odds of a diagnosis of sole ulcer increased substantially (2 and 3+ examinations had odds ratios (ORs) of 1.42 and 3.42 in TS herds and 2.77 and 6.89 in LH herds). Breed had a large effect on the risk of sole ulcer with Holsteins 2.89 times more likely to be affected than Ayrshires in TS herds and 2.94 times in LH herds. In TS herds, the presence of other hoof lesions such as haemorrhages (OR = 2.97), heel-horn erosions (OR = 2.10) and corkscrew claw (OR = 2.83) increased the risk of a sole ulcer developing. In LH herds, only haemorrhages (OR = 1.80) were a significant risk factor when parity was > or = 2. In TS herds, use of mats (compared to hard flooring) significantly reduced the risk of sole ulcers (OR = 0.49). The effect of parity on the risk of sole ulcer was greatest when parity > or = 4 but this effect was only significant in tie stalls (OR = 1.86). When analyses were restricted to cows with parity > or = 2, similar results were obtained for the risk factors identified above. In addition, parity became highly significant in TS and LH (OR 2.31 and 2.23, respectively when parity was 4+). In TS herds, herd average milk production was significantly associated with a decrease risk of sole ulcer (OR = 1.28 per 1000 kg decrease) but there was no effect of production at the cow level (measured as deviation from the herd mean). No significant effects of production were observed in LH herds.

Relationships among milk urea-nitrogen, dietary parameters, and fecal nitrogen in commercial dairy herds.

Our study objective was to determine the ability of milk urea-nitrogen concentrations ([MUN]) to predict fecal nitrogen concentrations ([Fecal N]) in commercial dairy herds. A total of 83 dairy herds were each visited 3 times within 48 h after a monthly herd milk test. For each farm visit, forages were sampled for nutrient analyses, which were entered into a computerized ration evaluator, and fecal samples were taken per rectum from each of 6 cows (2 early-, 2 mid-, and 2 late-lactation). Fecal samples were pooled, mixed, and analyzed for nitrogen content. Fecal nitrogen concentrations were compared with the routinely measured nutritional parameters from the ration evaluation, and the herd average [MUN] for the previous milk test date using mixed linear regression analyses. Total protein supplied in the ration was significantly positively associated with [Fecal N], but herd average [MUN] was not associated (P > 0.10) with [Fecal N].

Estimate of the direct production losses in Canadian dairy herds with subclinical Mycobacterium avium subspecies paratuberculosis infection.

The objective of this study was to estimate the annual losses from Mycobacterium avium subspecies paratuberculosis (MAP) for an average, MAP-seropositive, Canadian dairy herd. A partial-budget simulation model was developed with 4 components of direct production losses (decreased milk production, premature voluntary culling, mortality, and reproductive losses). Input values were obtained primarily from a national seroprevalence survey of 373 Canadian dairy farms in 8 of 10 provinces. The model took into account the variability and uncertainty of the required input values; consequently, it produced probability distributions of the estimated losses. For an average Canadian dairy herd with 12.7% of 61 cows seropositive for MAP, the mean loss was $2992 (95% C.I., $143 to $9741) annually, or $49 per cow per year. Additional culling, decreased milk production, mortality, and reproductive losses accounted for 46%, 9%, 16%, and 29% of the losses, respectively. Canadian dairy producers should use best management practices to reduce these substantial annual losses.

Quantitative epidemiology: progress and challenges.

This manuscript is derived from a presentation at the 2006 AVEPM--Schwabe Symposium which honoured the 2006 recipient of the Calvin Schwabe Award--Dr. S. Wayne Martin. Throughout his career, Dr. Martin was instrumental in furthering the development of quantitative epidemiology. This manuscript highlights some of the recent advances in quantitative methods used in veterinary epidemiology and identifies some of the challenges we still face in the use of these methods. Areas discussed are: methods for analyzing hierarchical data, Bayesian methods, survival analysis and integrating causal thinking and statistical analyses.

Factors affecting Canadian veterinarians' use of analgesics when dehorning beef and dairy calves.

Data collected through a national, randomized mail survey (response rate 50%) were used to identify reasons why veterinarians were likely (i) to use analgesic drugs when dehorning calves, and (ii) to perceive dehorning without analgesia as very painful. Logistic regression analysis indicated that veterinarians were more likely to be analgesic users the more they perceived that dehorning without analgesia was painful (OR = 1.7, P < 0.001). Other positive influences were if the veterinarian worked in British Columbia or Alberta (OR = 5.9, P = 0.005), and if they were primarily in dairy practice (OR = 3.7, P = 0.012) rather than beef practice. This effect of dairy practice was negated if the veterinarian also perceived that owners were unwilling to pay for analgesia (interaction term: OR = 0.25, P = 0.038). Veterinarians were also less likely to perceive dehorning without analgesia as very painful if they perceived that owners were unwilling to pay (OR = 0.58, P = 0.029). However, this effect on pain perception was offset by concern for personal safety (OR = 2.7, P = 0.015). The results are consistent with the relatively high level of outreach about animal welfare among farmers and veterinarians in the western provinces. The results confirm that many veterinarians' approach to pain management for dehorning is influenced considerably by concern about cost. However, pain management for dehorning is not expensive and there is unequivocal evidence that dehorning calves without pain management causes significant distress. Continuing education of veterinarians should help to increase analgesic usage.

Canadian veterinarians' use of analgesics in cattle, pigs, and horses in 2004 and 2005.

Anecdotal evidence suggests that many veterinarians may not use analgesics in livestock for routine surgical procedures or painful disease states. To investigate this, we conducted a national mail survey of a random sample of 1431 Canadian veterinarians (response rate, 50.1%). Questions primarily concerned veterinarians' analgesic usage for common surgeries and medical conditions in beef and dairy cattle, pigs, and horses, and attitudes toward pain management. More than 90% of veterinarians used analgesic drugs for equine surgeries, for cesarean section in sows and cows, and for bovine claw amputation and omentopexy. However, in these and other categories, the analgesics used were often inadequate, and many veterinarians did not give analgesics to young animals. When castrated, < 0.001% of piglets received analgesia, compared with 6.9% of beef calves and 18.7% of dairy calves < or = 6 mo of age, 19.9% of beef calves and 33.2% of dairy calves > 6 mo of age, and 95.8% of horses. Respondents largely agreed that there are no long-acting, cost-effective analgesics available for use in livestock (median rating 8/10; interquartile range 4-9), and that the long or unknown withdrawal periods of some drugs outweighed the benefits of using them (median rating 7/10; interquartile range 4-9). The results indicate an urgent need for veterinarians to manage pain in livestock better. Continuing education would help, as would an increase in the number of approved, cost-effective analgesic drugs with known withdrawal periods.