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.

Automated summary

The study examined temporal trends in BVD testing in New Zealand dairy herds and used historical data to predict herd-level BVD incursions. The results showed a decrease in the prevalence of herds with positive BVD test results over the study period, and identified herd size, previous PCR status, and change in S/P ratio as significant predictors for herd-level BVD incursions in the 2017 season. The predictive model had poor discrimination with an AUC of 0.695.