Social and ethical implications of data and technology use on farms: a qualitative study of Swedish dairy and pig farmers.

Introduction: Livestock farmers are being increasingly encouraged to adopt digital health technologies on their farms. Digital innovations may have unintended consequences, but there tends to be a pro-innovation bias in previous literature. This has led to a movement towards "responsible innovation," an approach that questions the social and ethical challenges of research and innovation. This paper explores the social and ethical issues of data and technologies on Swedish dairy and pig farms from a critical perspective. Methods: Six focus groups were conducted with thirteen dairy and thirteen pig farmers. The data were analysed using reflexive thematic analysis and a digital critical health lens, which focuses on concepts of identity and power. Results and discussion: The analysis generated four themes: extending the self, sense of agency, quantifying animals, and managing human labour. The findings suggest that technologies can change and form the identities of farmers, their workers, and animals by increasing the visibility of behaviours and bodies through data collection. Technologies can also facilitate techniques of power such as conforming to norms, hierarchical surveillance, and segregation of populations based on data. There were many contradictions in the way that technology was used on farms which suggests that farmers cannot be dichotomised into those who are opposed to and those that support adoption of technologies. Emotions and morality played an important role in the way animals were managed and technologies were used by farmers. Thus, when developing innovations, we need to consider users' feelings and attachments towards the technologies. Technologies have different impacts on farmers and farm workers which suggests that we need to ensure that we understand the perspectives of multiple user groups when developing innovations, including those that might be least empowered.

Impact of sensor data pre-processing strategies and selection of machine learning algorithm on the prediction of metritis events in dairy cattle.

With all the sensor data currently generated at high frequency in dairy farms, there is potential for earlier diagnosis of postpartum diseases compared with traditional monitoring methodologies. Our objectives were 1) to compare the impact of sensor data pre-processing on classifier performance by using multiple time windows before a given metritis event, while considering other cow-level factors and farm-scheduled activities; 2) to compare the performance of random forest (RF), k-nearest neighbors (k-NN), and support vector machine (SVM) classifiers at different decision thresholds using different number of past observations (time-lags) for the detection of behavioral patterns associated with changes in metritis scores; and 3) to compare classifier performance between each one of the five behaviors registered every hour by an ear-tag 3-axis accelerometer (CowManager, Agis Autimatisering, Harmelen, Netherlands). A total of 239 metritis events were created by comparing metritis scores between two consecutive clinical evaluations from cows that were retrospectively selected from a dataset containing sensor data and health information during the first 21 days postpartum from June 2014 to May 2017. Hourly sensor data classified by the accelerometer as either ruminating, eating, not active (including both standing or lying), and two different levels of activity (active and high activity) behaviors corresponding to the 3 days before each metritis event were aggregated every 24-, 12-, 6-, and 3-hour time windows. Multiple time-lags were also used to determine the optimal number of past observations needed for optimal classification. Similarly, different decision thresholds were compared in terms of model performance. Depending on the classifier, algorithm hyperparameters were optimized using grid search (RF, k-NN, SVM) and random search (RF). All behaviors changed throughout the study period and showed distinct daily patterns. From the three algorithms, RF had the highest F1 score followed by k-NN and SVM. Furthermore, sensor data aggregated every 6- or 12-h time windows had the best model performance at multiple time-lags. We concluded that the data from the first 3 days post-partum should be discarded when studying metritis, and either one of the five behaviors measured with CowManager could be used when predicting metritis when sensor data were aggregated every 6- or 12-hour time windows, and using time-lags corresponding to 2-3 days before a given event, depending on the time window used. This study shows how to maximize sensor data in their potential for disease prediction, enhancing the performance of algorithms used in machine learning.

Data workflows and visualization in support of surveillance practice.

The Swedish National Veterinary Institute (SVA) is working on implementing reusable and adaptable workflows for epidemiological analysis and dynamic report generation to improve disease surveillance. Important components of this work include: data access, development environment, computational resources and cloud-based management. The development environment relies on Git for code collaboration and version control and the R language for statistical computing and data visualization. The computational resources include both local and cloud-based systems, with automatic workflows managed in the cloud. The workflows are designed to be flexible and adaptable to changing data sources and stakeholder demands, with the ultimate goal to create a robust infrastructure for the delivery of actionable epidemiological information.

Reducing Calf Mortality in Ethiopia.

Morbidity and mortality of young stock present economic and production challenges to livestock producers globally. In Ethiopia, calf morbidity and mortality rates, particularly due to diarrhea and respiratory disease, are high, limiting production, incomes, and the ability of farmers to improve their livelihoods. In this paper, we present findings from the combined experience of the Young Stock Mortality Reduction Consortium, which conducted epidemiological and intervention testing in calves across three production systems. This innovative alliance identified Cryptosporidium parvum and E. Coli K99 as the most common causes of diarrhea in pastoral and peri-urban calves; Strongyloides spp. as the most common fecal parasite in mixed crop-livestock and peri-urban calves; and bovine adenovirus, parainfluenza virus-3, and bovine respiratory syncytial virus as the most common respiratory pathogens in peri-urban calves. Furthermore, by improving producer knowledge with respect to fundamental livestock husbandry, feeding, housing, and neonatal care practices, calf mortality risk across production systems was reduced by 31.4 to 71.4% compared to baseline (between 10.5 and 32.1%), whereas risk of diarrhea was reduced by 52.6-75.3% (baseline between 11.4 and 30.4%) and risk of respiratory disease was reduced by 23.6-80.8% (baseline between 3.3 and 16.3%). These findings have informed scaling strategies and can potentially contribute to improved livestock productivity and human livelihoods in Ethiopia.

Prevalence of bovine viral diarrhea virus antibodies and risk factors in dairy cattle in Gondar city, Northwest Ethiopia.

Bovine viral diarrhea (BVD) is a disease that affects cattle and other ruminants worldwide and causes considerable economic losses. A cross-sectional study was carried out between December 2017 and July 2018 with the aim to estimate the prevalence of Bovine Viral Diarrhea virus (BVDV) antibodies and to identify potential risk factors associated with the occurrence of the disease in dairy cattle in peri-urban areas of Gondar city, Northwest Ethiopia. A total of 339 serum samples obtained from randomly selected dairy cattle aged 6 months and older were assayed using a BVDV antibody competitive-Enzyme Linked Immunosorbent Assay (c-ELISA) kit. Descriptive statistics were used to estimate antibody prevalence of BVDV at animal and herd-level and logistic regression was used to identify potential risk factors. The study findings showed that the animal-level antibody prevalence of BVDV in the study area was 26.84 % (95 % CI: 22.1 %-31.6 %) and the herd-level seroprevalence was 68.3 % (95 % CI: 56.2 %-80.4 %). Logistic regression model demonstrated that age >2 years (OR = 4.75, 95 % CI: 2.20-10.26), herd size >11 (OR = 7.28, 95 % CI: 2.50-21.22), and poor farm hygiene (OR = 3.69, 95 % CI: 1.94-7.02), are potential risk factors associated with BVDV infection (P < 0.05). However, sex, faecal consistency and housing system were not associated with BVDV serostatus. The animal- and herd-level seroprevalence reports in Northwest Ethiopia can serve as a baseline finding for future BVD epidemiological investigations and to inform future control programs in the study region.

Ranking the risk of animal-to-human spillover for newly discovered viruses.

The death toll and economic loss resulting from the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic are stark reminders that we are vulnerable to zoonotic viral threats. Strategies are needed to identify and characterize animal viruses that pose the greatest risk of spillover and spread in humans and inform public health interventions. Using expert opinion and scientific evidence, we identified host, viral, and environmental risk factors contributing to zoonotic virus spillover and spread in humans. We then developed a risk ranking framework and interactive web tool, SpillOver, that estimates a risk score for wildlife-origin viruses, creating a comparative risk assessment of viruses with uncharacterized zoonotic spillover potential alongside those already known to be zoonotic. Using data from testing 509,721 samples from 74,635 animals as part of a virus discovery project and public records of virus detections around the world, we ranked the spillover potential of 887 wildlife viruses. Validating the risk assessment, the top 12 were known zoonotic viruses, including SARS-CoV-2. Several newly detected wildlife viruses ranked higher than known zoonotic viruses. Using a scientifically informed process, we capitalized on the recent wealth of virus discovery data to systematically identify and prioritize targets for investigation. The publicly accessible SpillOver platform can be used by policy makers and health scientists to inform research and public health interventions for prevention and rapid control of disease outbreaks. SpillOver is a living, interactive database that can be refined over time to continue to improve the quality and public availability of information on viral threats to human health.

Characterization of the Temporal Trends in the Rate of Cattle Carcass Condemnations in the US and Dynamic Modeling of the Condemnation Reasons in California With a Seasonal Component.

Based on the 2016 National Cattlemen's Beef Association statistics, the cattle inventory in the US reached 93.5 million head, from which 30.5 million were commercial slaughter in 2016. California ranked fourth among all the US states that raise cattle and calves, with 5.15 million head and approximately 1.18 million slaughtered animals per year. Approximately 0.5% of cattle carcasses in the US are condemned each year, which has an important economic impact on cattle producers.In this study, we first described and compared the temporal trends of cattle carcass condemnations in all the US states from Jan-2005 to Dec-2014. Then, we focused on the condemnation reasons with a seasonal component in California and used dynamic harmonic regression (DHR) models both to model (from Jan-2005 to Dec-2011) and predict (from Jan-2012 to Dec-2014) the carcass condemnations rate in different time horizons (3 to 12 months).Data consisted of daily reports of 35 condemnation reasons per cattle type reported in 684 federally inspected slaughterhouses in the US from Jan-2005 to Dec-2014 and the monthly slaughtered animals per cattle type per states. Almost 1.5 million carcasses were condemned in the US during the 10 year study period (Jan 2005-Dec 2014), and around 40% were associated with three condemnation reasons: malignant lymphoma, septicemia and pneumonia. In California, emaciation, eosinophilic myositis and malignant lymphoma were the only condemnation reasons presenting seasonality and, therefore, the only ones selected to be modeled using DHRs. The DHR models for Jan-2005 to Dec-2011 were able to correctly model the dynamics of the emaciation, malignant lymphoma and eosinophilic myositis condemnation rates with coefficient of determination ( Rt2 ) of 0.98, 0.87 and 0.78, respectively. The DHR models for Jan-2012 to Dec-2014 were able to predict the rate of condemned carcasses 3 month ahead of time with mean relative prediction error of 33, 11, and 38%, respectively. The systematic analysis of carcass condemnations and slaughter data in a more real-time fashion could be used to identify changes in carcass condemnation trends and more timely support the implementation of prevention and mitigation strategies that reduce the number of carcass condemnations in the US.

Monitoring of Schmallenberg virus in Spanish wild artiodactyls, 2006-2015.

Schmallenberg disease is an emerging disease that affects domestic and wild ruminants in Europe. An epidemiological survey was carried out to assess exposure to Schmallenberg virus (SBV) in wild artiodactyls in Spain between 2006 and 2015. A total of 1751 sera from wild artiodactyls, including 1066 red deer, 304 fallow deer, 192 mouflon, 109 wild boar, 49 roe deer and 31 Spanish ibex were tested for antibodies against SBV by ELISA and confirmed by virus neutralization test. SBV was not detected between the 2006/2007 and the 2010/2011 hunting seasons. Overall seroprevalence (including samples collected between the 2011/2012 and 2014/2015 hunting seasons) was 14.6% (160/1099; 95%CI: 12.7-16.6). Mean SBV seroprevalence was 13.3±2.6% in red deer, 23.9±4.2% in fallow deer, 16.4±6.1% in mouflon and 2.8±3.1% in wild boar. No antibodies against SBV were found in roe deer or Spanish ibex. The presence of SBV RNA was confirmed in three of 255 (1.2%) spleen samples from wild ruminants analysed by rRT-PCR. In a multivariate mixed-effects logistic regression model, the main risk factors associated with SBV seroprevalence were: species (fallow deer, red deer and mouflon), age (adults) and interactions between hunting areas of more than 1000 hectares and hunting season (2012/2013, 2013/2014 and 2014/2015). The hypothesis of endemic circulation of SBV in the last few years is supported by the detection of SBV RNA in animals sampled in 2011 and 2015, as well as antibodies detected at low level in juveniles in 2012, 2013 and 2014. The results indicate that SBV circulated in wild ruminant populations in Spain during the same period when the virus was first reported in northern Europe, and at least five months before the first case was officially reported in livestock in Spain.