Diagnostic test validation studies when there is a perfect reference standard.

Before tools became available to consider diagnostic test validation studies where a 'gold-standard' is not available, new diagnostic tests were compared to a reference standard assumed to be highly accurate if not perfect. This paper reviews such 'traditional' situations with examples and methods of study design and analysis. Three situations are described, two where a perfect reference is available for either positive or negative animals, and one where the reference is perfect for both. Thus, here the authors review circumstances to be considered when validating a diagnostic test with a credible reference standard. An appropriate study design requires an unbiased selection of animals from the population to which a new test will be applied. Examples for calculating sample size and data analysis are provided. Finally, the authors discuss situations where it may be appropriate to include influential variables ('covariates') in a diagnostic test validation study..

Avant la mise au point d'outils permettant de concevoir des études de validation pour des tests candidats en l'absence d'un étalon de référence à l'exactitude parfaite (référence absolue ou gold standard), les tests à valider étaient comparés à un étalon de référence censé présenter un niveau élevé d'exactitude, à défaut d'être parfait. Les auteurs décrivent ces situations « classiques ¼ en donnant quelques exemples et en précisant les méthodes employées pour la conception et l'analyse de ces études. Ils décrivent trois situations : dans deux d'entre elles, un étalon de référence parfait existe, soit pour les animaux positifs, soit pour les animaux négatifs ; dans la troisième situation, l'étalon de référence est parfait pour les deux catégories. Ainsi, les circonstances prises en compte ici sont celles de la validation d'un test diagnostique au moyen d'un réactif de référence crédible. Une conception d'étude appropriée passe par une sélection non biaisée des individus composant l'échantillon au sein de la population animale à laquelle sera appliqué le test. Les auteurs donnent quelques exemples de calcul de la taille de l'échantillon et d'analyse des données. Enfin, ils examinent les situations où il peut être opportun d'inclure des variables influentes (« covariance ¼) dans l'étude de validation d'un test diagnostique.

Antes de que apareciesen herramientas que permiten juzgar los estudios de validación de pruebas de diagnóstico en ausencia de un modelo o patrón de calibración, las nuevas pruebas de diagnóstico eran comparadas con un patrón de referencia considerado muy exacto, cuando no perfecto. Los autores se refieren aquí a este tipo de situaciones "clásicas", exponiendo ejemplos y métodos de diseño de estudios y análisis de los datos obtenidos. Los autores describen tres tipos de situación: cuando se dispone de una referencia perfecta para animales positivos; cuando se dispone de una referencia perfecta para animales negativos; y cuando la referencia es perfecta para ambos casos. Para cada tipo de situación examinan los aspectos que hay que tener en cuenta al validar una prueba de diagnóstico con un patrón de referencia fiable. El diseño adecuado de un estudio exige una selección no sesgada de los animales a partir de una población a la que vaya a aplicarse la nueva prueba. Los autores ofrecen ejemplos del modo de calcular el tamaño muestral y analizar los datos de un estudio. Por último, examinan situaciones en las que pueda ser conveniente incluir en el estudio de validación uno o más factores que puedan influir (covariables).

Introduction - Validation of tests for OIE-listed diseases as fit-for-purpose in a world of evolving diagnostic technologies.

The World Organisation for Animal Health (OIE) has made leading contributions to the discipline of test validation science by providing standards and guidelines that inform the test validation process in terrestrial and aquatic animals. The OIE Manual of Diagnostic Tests and Vaccines for Terrestrial Animals, and the Manual of Diagnostic Tests for Aquatic Animals describe the test validation pathway in the context of fitness for purpose, elaborate on the importance of diagnostic sensitivity (DSe) and specificity (DSp) as measures of test accuracy, and designate additional factors (e.g. test cost, laboratory throughput capacity and rapidity of test results) that influence choices of a single test over others or the inclusion of a new test in a diagnostic process that includes multiple tests. This paper provides examples of each of the six main testing purposes listed in the Terrestrial Manual and describes additional metrics such as ruggedness and robustness that should be included in the validation of point-of-care tests. Challenges associated with new diagnostic technologies and platforms are described. Validated tests with estimates of DSe and DSp are needed to measure confidence in test results for OIE-listed diseases, to facilitate risk assessments related to animal movement, to estimate true prevalence, and for certification of disease freedom and use in epidemiological (risk factor) studies.

L'Organisation mondiale de la santé animale (OIE) a apporté d'importantes contributions dans le domaine de la validation des tests en élaborant des normes et des lignes directrices qui informent sur le processus de validation des tests chez les animaux terrestres et aquatiques. Le Manuel des tests de diagnostic et des vaccins pour les animaux terrestres et le Manuel des tests de diagnostic pour les animaux aquatiques de l'OIE décrivent le processus de validation des tests dans le contexte de leur aptitude à l'emploi, expliquent l'importance de la sensibilité (DSe) et de la spécificité (DSp) diagnostiques pour mesurer l'exactitude des tests, et désignent d'autres facteurs (ex. coût des tests, capacité de traitement des laboratoires et rapidité d'obtention des résultats des tests) qui influencent le choix d'un test par rapport à un autre ou l'inclusion d'un nouveau test dans un processus de diagnostic composé de multiples tests. Le présent article fournit des exemples pour chacun des six principaux objectifs définis pour les tests figurant dans le Manuel terrestre et décrit des mesures supplémentaires, telle la robustesse (aussi bien interne qu'externe), qu'il conviendrait d'inclure dans la validation des tests au point d'intervention. Il aborde également les défis soulevés par les nouvelles technologies et plateformes de diagnostic. Des tests validés accompagnés d'estimations de la DSe et de la DSp sont nécessaires pour mesurer la fiabilité des résultats des tests pour les maladies listées par l'OIE, faciliter les évaluations des risques associés aux mouvements des animaux, estimer le véritable taux de prévalence et certifier l'absence de maladies ; ils sont également indispensables pour les études (des facteurs de risque) épidémiologiques.

La Organización Mundial de Sanidad Animal (OIE), con su labor de elaboración de normas y directrices que fundamentan el proceso de validación de pruebas para enfermedades de los animales terrestres y acuáticos, ha hecho aportaciones punteras a la disciplina científica que se ocupa de la validación de pruebas. En su Manual de las Pruebas de Diagnóstico y de las Vacunas para los Animales Terrestres y su Manual de las Pruebas de Diagnóstico para los Animales Acuáticos, la OIE describe el procedimiento de validación de pruebas en clave de idoneidad para determinados propósitos, ahonda en la importancia de la sensibilidad y la especificidad diagnósticas (DSe y DSp) como medidas de la exactitud de una prueba y señala otros factores (como el costo de la prueba, la productividad del laboratorio o la rapidez de los resultados) que también influyen en la elección de una determinada prueba por delante de otras o en la inclusión de una nueva prueba en un proceso de diagnóstico que entraña el uso de varias. Los autores ofrecen ejemplos de cada uno de los seis principales propósitos con las que puede utilizarse una prueba, según vienen enunciados en el Manual Terrestre, y describen otros parámetros que es preciso tener en cuenta a la hora de validar pruebas practicadas en el punto de consulta, como la robustez o también la solidez (ruggedness en inglés; llamada a veces «robustez interlaboratorios¼). También describen las dificultades ligadas a nuevas tecnologías y plataformas de diagnóstico. Se necesitan pruebas validadas y acompañadas de un cálculo de la DSe y la DSp para fines tan diversos e importantes como medir la confianza que merecen los resultados de pruebas para enfermedades inscritas en las listas de la OIE, facilitar la evaluación del riesgo ligado al desplazamiento de animales, estimar la prevalencia real, certificar la ausencia de enfermedad o realizar estudios epidemiológicos (factores de riesgo).

Temporal trends in bulk tank milk antibody ELISA and PCR test results for bovine viral diarrhoea in New Zealand pastoral dairy herds.

Aims: To describe temporal trends in bulk milk antibody ELISA and PCR testing for bovine viral diarrhoea (BVD) in New Zealand pastoral dairy herds and to assess the use of historical accession data to predict herd-level BVD incursions. Methods: Data on all diagnostic testing of bulk milk for BVD performed by the Livestock Improvement Corporation (Hamilton, NZ) over eight lactation seasons from 1 June 2010 to 31 May 2018 were analysed. This included anonymised herd identification, geographic location, herd size, sample collection date, sample to positive (S/P) ratio for antibody ELISA results, and cycle threshold values for PCR detecting viral RNA. Multivariable logistic regression was used to explore the relationship between historical accession data and the risk of herds having at least one positive bulk milk PCR test result in the 2017 season. Results: There were 156,034 bulk milk BVD diagnostic testing accessions for 10,495 uniquely identified dairy herds over the 8-season period. The prevalence of tested herds with at least one positive bulk milk PCR test result decreased from 14.6% (407/2,786) in the 2010 season to 5.6% (355/6,309) in the 2017 season with similarly marked declines in S/P ratios. In the 2017 season, 2,961/6,309 (46.9%) herds had S/P ratios greater than the 0.75 cut-off value indicating recent or active BVD virus transmission within the herd while 1,422/6,309 (22.5%) herds were classified as having negative or low S/P ratios. Herds that cleared BVD from the milking herd experienced a mean decline in S/P ratio of 0.11 units per year (min 0.05; max 0.18). In the multivariable analysis, the overall incidence risk of herds experiencing a BVD incursion in the 2017 season was 3.8% (146/3,848) and there were three significant predictors in the final model: herd size, PCR status in the 2014 season, and change in S/P ratio between the 2014 and 2015 seasons. The area under the receiver operating curve for the final model was 0.695 indicating poor discrimination. Conclusions and clinical relevance: The prevalence of dairy herds in New Zealand with positive bulk milk PCR test results and high S/P ratios has decreased over time, suggesting fewer herds are actively infected with BVD and that herd immunity may also be declining. Although monitoring trends in bulk milk test results provides useful information on changes in individual herd status, it is difficult to accurately predict when new incursions will occur and farmers should continue to maintain good biosecurity.

Vaccination practices for Leptospira spp. on New Zealand dairy farms.

AIMS: To describe leptospiral vaccination practices in dairy herds in New Zealand and evaluate conformity with best practice guidelines issued by the New Zealand Veterinary Association using data from a questionnaire administered by participating veterinary practices. METHODS: A cross-sectional study of 200 randomly selected dairy farms stratified by herd size and region throughout New Zealand was conducted from January to April 2016 to investigate leptospiral vaccination practices in dairy herds in New Zealand. Using a pre-tested questionnaire administered during a face-to-face interview, vaccination practice details such as vaccine types, time, and age of vaccination and whether vaccines were administered by veterinary or farm staff, were collected. RESULTS: Leptospiral vaccination programmes had been implemented on 199/200 (99.5 (95% CI = 97.2-99.9)%) farms, and on 178 (89.4%) of those, programmes had been running for ≥5 years. Most farmers used bivalent vaccines containing antigens for leptospiral serovars Pomona and Hardjo (144/179 (80.4%) in calves, 112/167 (60.7%) in heifers, and 112/163 (68.7%) in cows), rather than trivalent vaccines which also include antigens for L. interrogans serovar Copenhageni. In total, 123/200 (61.5%) of farmers purchased only vaccinated animals but 51/199 (25.6%) were unsure of the vaccination status of purchased cattle. Sixty-one percent (105/172) of farmers had other livestock on their farms and of them, 78/186 (42%) vaccinated some or all for Leptospira spp. Leptospiral vaccines were administered always or sometimes with other animal remedies on 30/190 (15.8%) and 91/190 (47.9%) of farms, respectively. Most farmers had not made changes to their vaccination programme in the previous 5 years. Timing of first vaccination of calves ranged from 2 weeks to 10 months of age, with 112/189 (59.3%) vaccinating by 4 months of age. Approximately half of the farms followed the best practice guideline for the timing of vaccinations for calves (high-risk farms; 67/162; 41.4%) heifers (72/165, 43.6%), and cows (171/184; 92.9%). CONCLUSIONS: The results of this survey suggest that there is almost universal adoption of leptospiral vaccination for dairy cattle in New Zealand. However, there remain areas for improvement regarding the proportion of farmers following best practice guidelines and refinement of vaccination programmes, particularly with respect to timing of vaccination in calves.

Serological study of Leptospira interrogans serovar Copenhageni and L. borgpetersenii serovars Tarassovi and Ballum in beef cattle, sheep and deer in New Zealand.

AIMS: To estimate animal-level seroprevalence of Leptospira interrogans serovar Copenhageni and L. borgpetersenii serovars Ballum and Tarassovi, in beef cattle, sheep and deer on New Zealand farms, and herd/flock-level seroprevalence of any serovar when existing same-sera data for serovars Hardjobovis and Pomona were included, and to determine associations between risk factors and animal-level seroprevalence. METHODS: Banked sera from sheep (n = 82), beef (n = 54) and deer (n = 62) herds/flocks (n = 3,878 animals) from seven regions were analysed using the microscopic agglutination test. Titres of ≥48 were designated positive. Herds/flocks were considered positive if either ≥1, ≥2 or ≥3 animals were positive. Existing same-sera data for serovars Hardjobovis and Pomona were included to establish farm-level any-serovar seropositivity. Factors associated with serological status were analysed using generalised estimating equations. RESULTS: Animal-level seroprevalence for serovars Ballum, Copenhageni, and Tarassovi, respectively, was 13.7 (95% CI = 11.7-16.0)%, 12.6 (95% CI = 10.6-14.7)% and 18.0 (95% CI = 15.7-20.5)% for beef cattle, 10.5 (95% CI = 9.0-12.1)%, 16.7 (95% CI = 14.9-18.6)% and 14.0 (95% CI = 12.4-15.8)% for sheep and 6.6 (95% CI = 5.3-8.2)%, 15.5 (95% CI = 13.5-17.7)% and 3.6 (95% CI = 2.7-4.8)% for deer, respectively. Herd/flock-level seroprevalence for Ballum was 86.6, 52.4 and 39.0% for sheep, 85.2, 52.7 and 33.3% for beef cattle and 50.8, 27.9 and 21.3% for deer at definitions ≥1, ≥2 and ≥3 seropositive animals per species, respectively. For Copenhageni, corresponding data were 95.1, 73.2 and 56.1% for sheep, 68.5, 48.2 and 29.6% for beef cattle and 73.8, 57.4 and 41.0% for deer, and for Tarassovi, 80.5, 59.7 and 45.1% for sheep, 83.3, 68.5 and 61.1% for beef cattle, and 42.6, 16.4 and 4.9% for deer. Seropositivity to all serovars was observed from all regions, with some differences in seroprevalence observed between species and regions, but not between islands. Combining with Hardjobovis and Pomona data, herd/flock-level seropositivity for all animal species and all five Leptospira serovars was 100% at definition ≥1 animal positive, and 97.5 and 96.3% for sheep flocks, 87.8 and 97.8% for beef cattle herds, and 89.3 and 75% for deer herds at ≥2 and ≥3 animals positive, respectively. CONCLUSIONS: Seropositivity to serovars Ballum, Copenhageni and Tarassovi is common in sheep, beef cattle and deer New Zealand and most, or all farms have ≥1 livestock species seropositive to ≥1 serovar. CLINICAL RELEVANCE: Serovars Ballum, Tarassovi and Copenhageni should be considered when clinical or subclinical signs of leptospirosis are observed in sheep, beef cattle or deer. Livestock sector workers are potentially at risk of exposure.

On-farm risk factors associated with Leptospira shedding in New Zealand dairy cattle.

This study aimed to evaluate risk factors associated with shedding of pathogenic Leptospira species in urine at animal and herd levels. In total, 200 dairy farms were randomly selected from the DairyNZ database. Urine samples were taken from 20 lactating, clinically normal cows in each herd between January and April 2016 and tested by real-time polymerase chain reaction (PCR) using gyrB as the target gene. Overall, 26.5% of 200 farms had at least one PCR positive cow and 2.4% of 4000 cows were shedding Leptospira in the urine. Using a questionnaire, information about risk factors at cow and farm level was collected via face-to-face interviews with farm owners and managers. Animals on all but one farm had been vaccinated against Hardjo and Pomona and cows on 54 of 200 (27%) farms had also been vaccinated against Copenhageni in at least one age group (calves, heifers and cows). Associations found to be statistically significant in univariate analysis (at P < 0.2) were assessed by multivariable logistic regression. Factors associated with shedding included cattle age (Odds ratio (OR) 0.82, 95% CI 0.71-0.95), keeping sheep (OR 5.57, 95% confidence interval (CI) 1.46-21.25) or dogs (OR 1.45, 95% CI 1.07-1.97) and managing milking cows in a single as opposed to multiple groups (OR 0.45, 95% CI 0.20-0.99). We conclude that younger cattle were more likely to be shedding Leptospira than older cattle and that the presence of sheep and dogs was associated with an increased risk of shedding in cows. Larger herds were at higher risk of having Leptospira shedders. However, none of the environmental risk factors that were assessed (e.g. access to standing water, drinking-water source), or wildlife abundance on-farm, or pasture were associated with shedding, possibly due to low statistical power, given the low overall shedding rate.

Effect of bismuth subnitrate on in vitro growth of major mastitis pathogens.

The mode of action of bismuth subnitrate in teat sealant formulations as a preventative for intramammary infections during the dry period is unknown. Although previous studies proposed an action mechanism-creating a physical barrier in the teat canal to prevent bacterial invasion-it has not been proven experimentally. We hypothesized that bismuth subnitrate has an inhibitory effect on bacterial growth, in addition to its barrier effect. The objective of this study was to assess the effect of bismuth subnitrate on bacterial growth of major mastitis-causing agents. A strain of Streptococcus uberis (SR115), 2 strains of Staphylococcus aureus (SA3971/59 and SA1), and a strain of Escherichia coli (P17.14291) were tested in vitro for their ability to grow in the presence or absence of bismuth subnitrate. Disk diffusion testing, impedance measurement, and evaluation of bacterial growth in shaking conditions were the methods used to test this hypothesis. A reduction of growth in the presence of bismuth subnitrate occurred for all the strains tested. However, we observed strain and species variations in the extent of growth inhibition. These results suggest that an inhibitory effect on bacterial growth by bismuth subnitrate could partially explain the efficacy of bismuth-based formulations for preventing intramammary infections over the dry period. Further research is required to test the effect of teat sealant formulations on bacterial growth.

Serological evidence for exposure to bovine viral diarrhoea virus in sheep co-grazed with beef cattle in New Zealand.

Aims: To determine whether sheep that co-grazed with cattle that were suspected to be positive for bovine viral diarrhoea (BVD) virus had serological evidence of exposure to the virus.Methods: Eighteen commercial farms that routinely co-grazed cattle and sheep in the same paddocks were recruited through purposive sampling. The recruiting veterinarians identified nine farms with cattle herds that were known or highly suspected to be positive for BVD and nine farms that were considered to be free of BVD. Blood samples were taken from 15 ewes aged 1 year on each farm and samples were submitted to a commercial diagnostic laboratory to test for antibodies against pestiviruses using an ELISA. All samples that were positive were then tested using a virus neutralisation test (VNT)for antibodies against BVD virus.Results: Of the 270 blood samples, 17 were positive for pestivirus antibodies by ELISA and these originated from two farms that were known or suspected to have BVD virus-positive cattle. None of the samples from the nine flocks co-grazed with cattle herds that were known or suspected to be BVD virus-negative were positive for pestivirus antibodies. Within the two positive farms, 2/15 samples from the first farm and 15/15 samples from the second farm were antibody-positive. When the 17 positive blood samples were submitted for VNT, all 15 samples from the second farm tested positive for BVD virus antibodies with the highest titre being 1:512.Conclusions and clinical relevance: In this small sample of New Zealand sheep and beef farms with suspected BVD infection in cattle, there was evidence of pestivirus exposure in co-grazed sheep. Although we were unable to confirm the origin of the exposure in these sheep, these findings highlight that farmers who are trying to eradicate BVD from their cattle should be mindful that the infection may also be circulating in sheep, and both populations should be considered a possible risk to each other for generating transient and persistent infections. Further work is needed to estimate the true prevalence of New Zealand sheep flocks that are affected by BVD and the associated economic impacts.

Practices and opinions of New Zealand beef cattle farmers towards bovine viral diarrhoea control in relation to real and perceived herd serological status.

Aims: To investigate the seroprevalence of infection with bovine viral diarrhoea (BVD) virus among 75 beef herds and seroconversion in cattle during early pregnancy, and to determine the practices and opinions of farmers towards BVD control and their association with real and perceived herd serological status.Methods: Blood samples were collected before mating in 75 beef herds across New Zealand from 15 unvaccinated heifers that had delivered their first calf that season. Serum samples were tested for BVD antibodies using ELISA individually, and after pooling samples for each farm. Animals that were antibody-negative were retested at either pregnancy diagnosis or weaning. Farmers were asked to complete a detailed survey about herd demographics, BVD testing and vaccination practices, and opinions towards national BVD control.Results: Based on the pooled serum antibody ELISA results, there were 28/75 (37%) negative herds, 15/75 (20%) suspect herds, and 32/75 (43%) positive herds. Of 1,117 animals sampled 729 (65.3%) tested negative for BVD virus antibodies; when retested, 47/589 (8.0%) animals from 13/55 (24%) herds had seroconverted. Among 71 famers providing survey responses 11 (15%) believed their herd was infected with BVD, 24 (34%) were unsure and 36 (51%) did not think their herd was infected. Only 19/71 (18%) farmers had performed any BVD testing within the past 5 years and 50/70 (71%) had not vaccinated any cattle for BVD. Support for national BVD eradication programme was strong in 51/71 (56%) respondents, but the biggest challenge to BVD control was considered to be famer compliance. Compared to farmers who did not think their herd was infected, more farmers who thought BVD was present in their herds had previously tested for BVD, would consider testing all replacement calves, and would support establishing a national BVD database; fewer would consider purchasing BVD tested or vaccinated cattle only.Conclusions and clinical relevance: Only 15% of the beef farmers in this study believed their herds were infected with BVD virus and few of them had undertaken BVD screening. Nevertheless many were supportive of implementing a national BVD control programme. It is likely that the lack of farmer awareness around BVD and the failure of farmers to recognise the potential impacts in their herds are hindering progress in controlling the disease in New Zealand. There are opportunities for New Zealand veterinarians to be more proactive in helping beef farmers explore BVD management options.

Assessing the use of diagnostic laboratory accession data to support national bovine viral diarrhoea control in New Zealand.

Aims: To assess the suitability of using existing national diagnostic laboratory testing data to support national bovine viral diarrhoea (BVD) research, surveillance, and control in New Zealand. Methods: Data on laboratory accessions for BVD diagnostic testing in New Zealand from 1 January 2015 to 31 December 2017 were provided by four commercial veterinary diagnostic companies. The data were integrated into a single dataset containing the unique accession number, sample submission date, farm location (territorial authority level), test type (bulk milk antibody-ELISA, bulk milk PCR, serum antibody-ELISA, blood/serum/tissue antigen-ELISA, or blood/serum/tissue PCR), and test results. Estimates for the number of registered cattle farms in each territorial authority were generated from the National Animal Identification and Tracing database. Results were summarised for July 2015 to June 2016 and July 2016 to June 2017. Results: There was a total of 59,007 unique BVD diagnostic test accessions including 39,920 (67.6%) for bulk milk antibody-ELISA, 27,832 (47.2%) for bulk milk PCR, 3,229 (5.5%) for serum antibody-ELISA, 9,132 (15.5%) for blood/serum/tissue antigen-ELISA, and 7,122 (12.1%) for blood/serum/tissue PCR. Of the 17,946 accessions for blood/serum/tissue samples, 4,316 (24.0%) were missing the herd production type and 6,678 (37.2%) were missing the animals age. Approximately 7,000/10,958 (65%) dairy herds and 1,600/43,611 (4%) beef herds were conducting annual BVD screening tests. In 2016/2017, the prevalence of accessions with ≥1 BVD-positive result was 40.6% for bulk milk antibody, 6.4% for bulk milk PCR, 45.6% for serum antibody, and 9.8% for blood/serum/tissue antigen-ELISA or PCR tests. There was substantial regional variation in both the percentage of herds testing for BVD and the prevalence of positive accessions. Following pooled serum antibody-ELISA, only 175/604 (29.0%) beef herds and 177/566 (31.3%) dairy herds had recorded follow-up testing. Conclusions and Clinical Relevance: Laboratory diagnostic accession data has the potential to provide valuable insights about BVD epidemiology in New Zealand, but there are significant limitations in the data collected and discrepancies in the different systems that each laboratory uses to measure, interpret, and record diagnostic data. There is a strong need to develop a more consistent national system for recording and sharing BVD test results to support BVD management at farm and industry levels. Abbreviations: BVD: Bovine viral diarrhoea; Ct: Cycle threshold; NAIT: National Animal Identification and Tracing; NZVP: New Zealand Veterinary Pathology; PI: Persistently infected; S/P: Sample to positive control.

Recommendations for the testing and control of bovine viral diarrhoea in New Zealand pastoral cattle production systems.

Eradicating bovine viral diarrhoea (BVD) from cattle populations requires a clear approach for determining the epidemiological status of individual herds and implementing the appropriate control measures to ensure the transmission cycle is cost-effectively broken. This is particularly important in countries such as New Zealand where there is currently no coordinated national programme and the herd-level decisions to control BVD are left to the discretion of individual farmers and veterinarians. To ensure greater consistency in the information being delivered by different stakeholders, we review the epidemiology of BVD in the context of New Zealand pastoral production systems and provides a series of simplified recommendations for the future control of BVD in beef and dairy herds. Based on analysis of BVD test accession data from commercial diagnostic laboratories, it has been estimated that 40.6% of dairy herds and 45.6% of beef herds tested had positive results for antibodies to BVD virus. While BVD continues to remain widespread and under voluntary control in New Zealand, it is recommended that herds test all individual mixed-age cows and replacement heifers for BVD virus or antigen and remove persistently infected animals from the breeding population. All new breeding animals that have entered the herd either through purchase or birth should also be tested for BVD virus. Biosecurity risks should be managed by reducing contacts with other herds and implementing targeted vaccination programmes. All individual purchased cattle should be tested and confirmed negative for BVD virus before being moved onto the buyer's property, even if the herd of origin had a negative antibody-based screening test. Herds should continue annual antigen or virus testing of all calves as soon as possible after birth to identify any persistently infected animals.

Control of clinical paratuberculosis in New Zealand pastoral livestock.

This review summarises current control measures for clinical paratuberculosis (Johne's disease; JD) in New Zealand pastoral livestock. Most New Zealand sheep, deer, beef and dairy cattle herds and flocks are infected by Mycobacterium avium ssp. paratuberculosis (Map). Dairy cattle and deer are mostly infected with bovine (Type II), and sheep and beef cattle with ovine (Type I) strains. Control in all industries is voluntary. While control in sheep and beef cattle is ad hoc, the dairy and deer industries have developed resources to assist development of farm-specific programmes. The primary target for all livestock is reduction of the incidence rate of clinical disease rather than bacterial eradication per se. For dairy farms, a nationally instituted JD-specific programme provides guidelines for risk management, monitoring and testing clinically suspect animals. While there is no formal programme for sheep farms, for those with annual prevalences of clinical disease >2%, especially fine wool breeds, vaccination may be a cost effective control option. The deer industry proactively monitors infection by a national abattoir surveillance programme and farmers with an apparent high disease incidence are encouraged to engage with a national network of trained consultants for management and control advice. Evaluation of the biological and economic effectiveness of control in all industries remains to be undertaken. Nevertheless, opportunities exist for farmers, who perceive significant JD problems in their herds/flocks, to participate in systematic best-practice activities that are likely to reduce the number of clinical infections with Map on their farms, and therefore the overall prevalence of JD in New Zealand's farming industries.

Development of an indirect ELISA for detection of antibody to wobbly possum disease virus in archival sera of Australian brushtail possums (Trichosurus vulpecula) in New Zealand.

AIMS: To develop an indirect ELISA based on recombinant nucleocapsid (rN) protein of wobbly possum disease (WPD) virus for investigation of the presence of WPD virus in Australian brushtail possums (Trichosurus vulpecula) in New Zealand. METHODS: Pre- and post-infection sera (n=15 and 16, respectively) obtained from a previous experimental challenge study were used for ELISA development. Sera were characterised as positive or negative for antibody to WPD virus based on western-blot using WPD virus rN protein as antigen. An additional 215 archival serum samples, collected between 2000-2016 from five different regions of New Zealand, were also tested using the ELISA. Bayesian modelling of corrected optical density at 450 nm (OD450) results from the ELISA was used to obtain estimates of receiver operating characteristic (ROC) curves to establish cut-off values for the ELISA, and to estimate the prevalence of antibody to WPD virus. RESULTS: Western blot analysis showed 5/14 (36%) pre-infection sera and 11/11 (100%) post-infection sera from experimentally infected possums were positive for antibodies to WPD virus. Bayesian estimates of the ROC curves established cut-off values of OD450≥0.41 for samples positive, and OD450<0.28 for samples negative for antibody to WPD virus, for sera diluted 1:100 for the ELISA. Based on the model, the estimated proportion of samples with antibodies to WPD virus was 0.30 (95% probability interval=0.196-0.418). Of the 230 archival serum samples tested using the ELISA, 48 (20.9%) were positive for antibody to WPD virus, 155 (67.4%) were negative and 27 (11.7%) equivocal, using the established cut-off values. The proportion of samples positive for WPD virus antibody differed between geographical regions (p<0.001). CONCLUSION: The results suggested that WPD virus or a related virus has circulated among possums in New Zealand with differences in the proportion of antibody-positive samples from different geographical regions. Antibodies to WPD virus did not seem to protect possums from disease following experimental infection, as one third of possums from the previous challenge study showed evidence of pre-existing antibody at the time of challenge. These results provide further support for existence of different pathotypes of WPD virus, but the exact determinants of protection against WPD and epidemiology of infection in various regions of New Zealand remain to be established. CLINICAL RELEVANCE: Availability of the indirect ELISA for detection of WPD virus antibody will facilitate prospective epidemiological investigation of WPD virus circulation in wild possum populations in New Zealand.

Seroprevalence and herd-level risk factors for seroprevalence of Leptospira spp. in sheep, beef cattle and deer in New Zealand.

AIMS To determine seroprevalence of Leptospira borgpetersenii serovar Hardjo and L. interrogans serovar Pomona in beef cattle, sheep and deer in New Zealand and the association between farm-level risk factors and seroprevalence. METHODS Between June 2009 and July 2010, 20 serum samples per flock or herd were collected from 162 sheep flocks, and 116 beef cattle and 99 deer herds from 238 farms, along with farm data by interview. Samples were tested for antibodies to serovars Hardjo and Pomona by microscopic agglutination testing, with a titre ≥48 being positive. Species-specific associations between herd-level seroprevalence (number of seropositive animals, for each serovar, divided by the number of animals tested) and herd-level risk factors were determined by multivariable logistic regression analysis. Vaccinated animals were excluded from seroprevalence estimates but included in multivariable analyses. RESULTS For sheep (n=3,339), animal-level seroprevalence was 43.6 (95% CI=41.9-45.3)% for serovar Hardjo and 14.1 (95% CI=12.9-15.3)% for serovar Pomona; for beef cattle(n=1,886), it was 45.6 (95% CI=43.3-47.9)% for Hardjo and 19.6 (95% CI=17.9-21.5)% for Pomona; and for deer (n=1,870), it was 26.3 (95% CI=24.3-28.4)% for Hardjo, 8.8 (95% CI=7.6-10.2)% for Pomona. In sheep flocks (n=161), flock-level prevalence for Hardjo varied from 77.9-91.3%, and for Pomona from 40.4-73.9%, when ≥1, ≥2 or ≥3 animals were seropositive. In beef herds (n=95), herd-level prevalence for Hardjo varied from 79.0-90.5%, and for Pomona from 42.1-68.4%. In deer herds (n=93), herd-level prevalence for Hardjo varied from 45.2-59.1%, and for Pomona from 22.6-48.4%. For sheep flocks, herd-level seroprevalence for Hardjo was associated with flock size (OR=1.56) and number of dogs (OR=0.75), and for Pomona, seroprevalence varied with region. For beef cattle, herd-level seroprevalence for Hardjo was associated with herd size (OR=1.4), presence of dams (OR=0.6) and vaccination (OR=2.9), and for Pomona, co-grazing with deer (OR=0.4), vaccination (OR=3.22), presence of dams (OR=0.2) and streams (OR=2.7). For deer herds, seroprevalence for Hardjo or Pomona was associated with herd size (OR=1.6 and 1.8) and varied with region, and for Pomona seroprevalence varied with season (summer vs. winter: OR=4.8). CONCLUSIONS Serovars Hardjo and Pomona were highly prevalent at herd and animal levels, with serovar Hardjo highest in all species. Larger herd size was the common risk factor for seroprevalence in all livestock species.

Elimination of bovine viral diarrhoea virus in New Zealand: a review of research progress and future directions.

The major impacts of bovine viral diarrhoea (BVD) on cattle health and production have prompted many countries to embark on national elimination programmes. These programmes typically involve identifying and removing persistently infected (PI) cattle in infected herds and implementing biosecurity measures, such as pre- or post-movement testing. In order to design a systematic national control programme to eliminate BVD in New Zealand, which achieves the greatest benefits to the industries at the lowest cost to individual farmers, an accurate understanding is necessary of the epidemiology, economics and social motivation for BVD control in New Zealand. In this article we briefly review the pathogenesis of BVD, transmission and diagnosis of BVD virus infection, and effectiveness of vaccination. We summarise the current state of knowledge of the prevalence, risk factors for transmission, and financial impacts of BVD in New Zealand. We describe control programmes in Europe and then discuss the challenges that must be addressed to design a cost-effective national control programme to eliminate BVD in New Zealand.

Associations between liveweight, body condition score and previous reproductive outcomes, and the risk of ewes bred at 18-months of age being dry at docking.

AIMS Firstly, to investigate associations between liveweight and body condition score (BCS) of two-tooth ewes (18-months-old at breeding) at breeding, pregnancy diagnosis (PD) and pre-lambing and the risk of being dry at docking, on commercial New Zealand sheep farms. Secondly, to investigate the association between previous reproductive outcomes as ewe lambs, and risk of being dry at docking as two-tooth ewes. METHODS Two-tooth ewes (n=9,006) were enrolled in four cohorts from three commercial sheep farms between 2010-14. Ewes were weighed and BCS assessed immediately pre-breeding, at PD (mid-pregnancy) and pre-lambing. At PD, ewes were identified as either non-pregnant, or having single or multiple fetuses. Palpation and examination of udders at docking was used to classify each ewe as either lactating or dry at docking. RESULTS Overall, 437/8,025 (5.4%) of ewes that were diagnosed pregnant at PD were dry at docking. The risk of being dry at docking decreased with increasing pre-lambing conceptus adjusted liveweight (CALW) on all farms (p≤0.002); for 2010-born ewes from Farm A the OR=0.87 (95% CI=0.81-0.92); for Farm B the OR=0.88 (95% CI=0.83-0.92) and for Farm C the OR=0.86 (95% CI=0.79-0.95). The risk of being dry at docking also decreased with increasing CALW gain from PD to pre-lambing for all farms (p≤0.003); for 2010-born ewes from Farm A the OR=0.89 (95% CI=0.84-0.94); for Farm B the OR=0.85 (95% CI=0.81-0.89) and for Farm C the OR=0.88 (95% CI=0.80-0.96). There was no association between BCS at breeding, PD or pre-lambing and the risk of being dry at docking for 2010-born ewes from Farm A, Farm B or Farm C (p>0.05). For 2010-born ewes on Farm A, the risk of being dry at docking was greater for two-tooth ewes that were previously dry at docking as ewe lambs than those that were lactating at docking as ewe lambs (OR=1.7 (95% CI=1.1-2.8); p=0.018), but this difference was not observed for ewes on Farm B or Farm C (p>0.5). CONCLUSIONS There were negative associations between ewe CALW pre-lambing, and CALW gain between PD and pre-lambing, and risk of being dry at docking. For all cohorts, heavier ewes and those that gained CALW were less likely to be dry at docking than lighter ewes or those that lost CALW, however these relationships varied between cohorts.

Invited review: A systematic literature review and meta-analysis of mortality and culling in dairy cattle.

Dairy industries and individual farmers are concerned about mortality and culling of dairy animals. This is because the timing and fates of animals that exit dairy farms have important animal welfare and economic consequences that reflect the conditions under which they are farmed and the efficiency of their production systems. Reports from a few countries have indicated increased incidence of mortality, and occasionally culling, of dairy animals in recent decades, and these changes have been associated with intensification of production systems. Dairy industries and farmers need benchmarks for culling and mortality against which they can compare themselves, as well as improved understanding of the extent of any change and of any associated factors. We reasoned that a systematic literature review and meta-analysis of scientific articles published between 1989 and 2014 would allow us to determine whether these reports were universal, to quantify any change over time, and to investigate whether production systems or study factors were associated with culling and mortality. From 3,275 articles retrieved from databases and manual searching of cited articles, 118 articles were appraised independently by 2 assessors, and 51 articles representing 54 studies were determined to be eligible for review and meta-analysis. We estimated that both the annual incidence risk (IR) and incidence density of mortality of cows had increased significantly from 0.02 per cow and 2.32 per 100 cow-years, to 0.04 per cow and 3.75 per 100 cow-years, an increase per decade of 0.02 per cow and 1.42 per 100 cow-years, respectively. We also estimated that the annual IR of culling attributed to low production had declined significantly from 0.07 to 0.05 and that the IR of perinatal, but not neonatal, mortality had increased significantly from 0.04 to 0.06 per decade. We found no evidence of change in overall annual IR of culling of cows over time or any association between study design factors and the IR or incidence density of culling or mortality. These findings provide benchmarks for describing culling and mortality, and should encourage farmers and researchers in countries with modern dairy industries to discover and implement management strategies to reduce the animal welfare and economic costs associated with these changes.

Farm and cow-level prevalence of bovine digital dermatitis on dairy farms in Taranaki, New Zealand.

AIMS: The aims of this cross-sectional study were to investigate the herd and cow-level prevalence of bovine digital dermatitis (BDD) in dairy farms in the northern Taranaki region of New Zealand, and to identify whether there was any spatial clustering of herds with the disease. METHODS: A survey of 224 dairy farms in the northern Taranaki region of New Zealand was undertaken from September 2014 to February 2015. Following training in robust criteria to confirm BDD visually, a technician inspected the rear feet of every milking cow on the farms during milking. The identity of cows with lesions and the feet involved were recorded. The proportion of cows affected among the inspected population (cow-level prevalence), the proportion of a herd affected (farm-level prevalence), and proportion of farms with ≥1 cow with lesions, were calculated. A bivariate K function analysis was then used to assess whether farms with ≥1 cow with lesions were clustered, after accounting for the distribution of the farms involved in the study. RESULTS: Bovine digital dermatitis lesions were observed on 143/224 (63.8 (95% CI=57.5-70.1)%) farms. Within-farm prevalence was 0% on 81 (36.2%) farms, between >0 and <3% on 120 (53.5%) farms, with a maximum prevalence of 12.7% on one farm. Overall, cow-level prevalence was 707/60,455 (1.2 (95% CI=0.9-3.0)%), and on affected farms was 707/41,116 (1.7 (95% CI=1.4-2.1)%). In affected cows, 268/707 (37.9%) had a lesion on left foot only, 262/707 (37.1%) on the right foot only and 177/707 (25.0%) on both feet. The K function analysis showed no evidence of clustering of farms with BDD. CONCLUSIONS: Bovine digital dermatitis was widespread among the survey farms, but there was no evidence that there was any clustering of herds with BDD. The cow-level prevalence on affected farms was much lower than reported elsewhere. CLINICAL RELEVANCE: Although the prevalence at the cow level was low, if these data are representative of other regions of New Zealand, BDD could easily become a major problem on dairy farms in New Zealand, as has been observed in other countries.

Technical note: The role of circulating low-density lipoprotein levels as a phenotypic marker for Holstein cholesterol deficiency in dairy cattle.

With the recent discovery of a Holstein cholesterol deficiency (HCD) haplotype, the USDA has labeled many dairy animals as HCD carriers based on haplotype and pedigree analysis. We set out to investigate the effect of HCD status on various cholesterol transport molecules, namely low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol, and triglycerides in both males and females. A genome-wide association study was also conducted to narrow down the genomic region correlated with varying LDL-C levels. In the study, 34 HCD carrier animals showed significantly lower cholesterol and LDL-C levels compared with their 34 closely related, non-HCD controls. The genome-wide association study based on 73 animals using 56,198 SNP markers revealed an association with chromosome 11 in the region of 66,218,925 to 66,946,746 bp. We also tested the effect of HCD status on sperm quality traits using fresh ejaculates and frozen-thawed semen samples, but did not find any discriminating effects. Our study has demonstrated the use of LDL-C as a key phenotypic marker for determining HCD status in dairy cattle and this is the first study that clearly shows a cause-effect relationship of the HCD haplotype on circulating LDL-C.

Risk of infection and associated influenza-like disease among abattoir workers due to two Leptospira species.

The aims of this study were to determine the annual incidence of infection with Leptospira interrogans serovar Pomona and/or Leptospira borgpetersenii serovar Hardjo and its association with influenza-like illness (ILI) in meat workers in New Zealand. Sera were collected twice, 50-61 weeks apart, from 592 workers at eight abattoirs slaughtering sheep (n = 4), cattle (n = 2) and deer (n = 2), and tested by the microscopic agglutination test for Hardjo and Pomona. Forty-nine (8·3%) participants either seroconverted or had at least a twofold increased serological titre against either serovar. The worker infection risk was higher in sheep abattoirs (11·9%) than in abattoirs processing deer (0%) or cattle (1·2%) (P < 0·01). The annualized risk of mild (ILI) or severe clinical disease attributable to the two Leptospira serovars was 2·7%. This study has demonstrated that meat workers are at substantial risk of infection and clinical disease, suggesting further investigation of infection sources and preventive measures are warranted.