Monitoring udder health on routinely collected census data: Evaluating the short- to mid-term consequences of implementing selective dry cow treatment

In 2013, the preventive use of antimicrobials in Dutch livestock was prohibited, including a ban on the blanket application of antimicrobial dry cow treatment (BDCT). Since then, selective dry cow treatment (SDCT) has become the standard approach. In this study, we aimed to determine the effect of the ban on BDCT and the extent of the subsequent adoption of SDCT on antimicrobial usage (AMU) and udder health on Dutch dairy farms. In the Dutch cattle health surveillance system, AMU for dry cow treatment (AMU_DCT), AMU for intramammary treatment at any point in time (AMU_IMM), and udder health indicators are routinely and continuously monitored. This provided the opportunity to study associations among SDCT, udder health, and AMU on census data of approximately 17,000 dairy herds, with about 1.67 million cows in total (>2 yr old) at one moment in time in the period from 2013 until 2017. Six udder health parameters were evaluated using multivariable population-averaged generalized estimating equation models. The year in which the ban on BDCT was introduced (2013) was compared with the period thereafter (2014-2017). Additionally, AMU_IMM and AMU_DCT were included as independent variables to evaluate whether the extent to which SDCT was implemented on the herd level was associated with udder health. Demographic parameters were included as potential confounders. Since the ban on BDCT, overall declines of 63% in AMU_DCT and 15% in AMU_IMM were observed. The raw data show an improvement in 5 out of 6 evaluated udder health parameters between 2013 and 2017. Nevertheless, the multivariable model results showed that the period since the ban on BDCT was associated with a small but significant increase in the percentage of cows with high somatic cell count (HSCC) and new HSCC (+0.41% and +0.06%, respectively). Additionally, the probability of belonging to the group of herds with more than 25% of primiparous cows having HSCC during the start of lactation increased slightly, associated with the period after which BDCT was banned (odds ratio = 1.08). The probability of belonging to the group of herds with more than 25% cows having a persistent HSCC during the dry period was not affected and bulk milk somatic cell count showed a slight but significant reduction. The only udder health parameter that notably worsened during the study period was the probability of belonging to the group of herds with more than 25% of multiparous cows with a new HSCC after the dry period, during the start of lactation (odds ratio = 1.23). In herds where the farmer decided not to apply any dry cow therapy (≈20% of all herds), all udder health parameters were poorer compared with herds in which dry cow therapy was applied to some extent. The ban on BDCT and implementation of SDCT in the Netherlands was associated with a considerable reduction in AMU without a major impairment in udder health at the national level. Although negative effects of changed dry cow management were observed in some herds, we conclude that SDCT can be introduced without substantial negative effects on udder health.