This study measures the dynamic technical and udder health management inefficiencies of a sample of Wisconsin dairy farms. Udder health management inefficiency is defined as a farmer\'s failure to achieve lower levels of milk somatic cell counts compared with those of the best-practice farmers within the sample. The study proposes the treatment of somatic cell count as an undesirable output. We measured inefficiency using a dynamic directional distance function that accounts simultaneously for the expansion of desirable outputs and investments in capital assets, and contraction of undesirable output and variable inputs. In a second step, a bootstrap truncated regression was used to analyze factors that cause differences in dynamic technical and udder health management inefficiencies. Results showed that the sample farmers had considerably higher udder health management inefficiency scores than technical inefficiency scores. The results of the second-stage analysis indicated that technical inefficiency was influenced by summer precipitation and farmers\' financial characteristics, and was regionally heterogeneous. Udder health management inefficiency was affected by summer temperature and nonfarm income. By ranking farms in this study in terms of technical and udder health management inefficiency, we allowed inefficient farms to compare their performance with that of their efficient peers, and thus identify targets for production and udder health management improvement efforts. Finally, although our study focused on farmers\' performances with respect to udder health management, the proposed modeling framework can be applied to the management of other animal diseases and welfare conditions.

Our objective was to evaluate a 200,000 cells/mL somatic cell count (SCC) cut-point on both the quarter and composite level to determine its effectiveness at identifying subclinical mastitis infections in one commercial dairy herd in Central New York. Milk samples from 107 Holstein cows were used for analysis. All cows were eligible for enrollment provided they had 4 working udder quarters, were >14 and <365 d in milk, and had no clinical mastitis event or treatment with intramammary antibiotics ≤14 d prior to d of sampling. A total of 428 quarter and 107 composite samples from 34 primiparous and 73 multiparous animals were analyzed for total SCC and aerobic culture. Performance of SCC for identification of subclinically infected animals was evaluated against the gold standard aerobic culture. Sensitivities for a 200,000 cells/mL cut-point on both the quarter and composite basis were 28.6%, and specificities were 91.5% and 87.3% on the quarter and composite basis, respectively. Receiver operating characteristic (ROC) curves determined on a quarter basis found the cut-point that optimized the sensitivity and specificity of a positive culture was 32,000 cells/mL, with a sensitivity of 76.2%, a specificity of 62.4%, and an area under the curve (AUC) of 0.73. The ROC curve cut-point that optimized the sensitivity and specificity for the composite samples was 75,000 cells/mL, with a sensitivity of 57.1%, a specificity of 78.9%, and an AUC of 0.67. A large proportion of culture positive primiparous cows (38%) had low SCC (median of 101,000 cells/mL on the quarter and 80,000 cells/mL on the composite level), and therefore, when multiparous cows were examined separately, the cut-point that optimized sensitivity and specificity on the quarter basis increased to 645,000 cells/mL with a corresponding sensitivity of 34.8%, specificity of 97.5%, and AUC of 0.65. On the composite basis, the cut-point based on multiparous cows only was 152,000 cells/mL, with a corresponding sensitivity of 60.0%, and specificity of 82.0%, and an AUC of 0.65. Our data indicate that the 200,000 cells/mL cut-point was inefficient in identifying subclinically infected animals, regardless of whether quarter or composite sampling was used. The low prevalence of subclinical infections as well as the large proportion of minor pathogens, especially in primiparous animals, contributed to this inefficiency. This case study provides evidence that, with continued improvement upon mastitis control and reduction in major mastitis pathogens, blanket cut-points may no longer provide the same diagnostic usefulness as they once did.

An obligatory eradication programme for Bovine Virus Diarrhoea (BVD) was implemented in Switzerland in 2008. Between 2008 and 2012, all bovines were tested for antigen or antibodies against BVDV. By the year 2012, eradication was completed in the majority of farms. A decrease of the prevalence of persistently infected (PI) newborn calves was observed from 1.4% in 2008 to <0.02% in 2012. The objective of the present study was to assess the effects of BVD eradication on different parameters of animal health, production and fertility in Swiss dairy herds which had completed the eradication programme. A matched case-control study was carried out using data from two periods, before (Period 1) and after (Period 2) the active phase of eradication. Case farms had at least two PI animals detected before or during the eradication; controls were BVD-free and matched for region, herd size and use of alpine pasture. A total of 110 farmers (55 pairs) were recruited. During a phone interview, a questionnaire about farm characteristics, animal health and appreciation of the BVD eradication programme was filled in. Breeding data and milk test day records were also analyzed. Parameters were first compared between (i) case and control herds before eradication, and (ii) Period 1 and Period 2 for case herds only. Milk yield (MY), bulk milk somatic cell count (BMSCC), prevalence of subclinical mastitis (SCM), and non-return rate (NRR) showed a p-value<0.25 in at least one of the univariable comparisons and were thus further analyzed with a multilevel mixed-effects model to account for repeated measures over time. In order to assess whether changes in health status over time were due to BVD eradication, an interaction variable between period and group (case-control) was created (IA). Except for MY, the IA was significant for all parameters modelled. Despite an overall p-value of 0.27, case herds tended to have a higher MY after eradication (β=0.53, p=0.050). For BMSCC and SCM, case herds had higher values than controls in both periods; udder health was significantly improved in control herds and it remained stable in case herds, with a slight decrease of BMSCC (β=-0.19, p=0.010). Finally, among fertility parameters, NRR showed a general improvement but it was significant only in control herds (β=0.29, p=0.019). Even though the effects of the eradication programme measured in this study were less pronounced than expected, 73% of the participants of this study had a positive attitude towards the campaign.