Abstract
Increased consumption of dairy products together with the increased growth in global population means there will be a demand for approximately 600 billion kilograms more milk worldwide in 2067 compared to today. The UK is a net importer of dairy products and recent trends show an increase in dairy imports, including the importation of raw milk . The UK dairy industry is facing huge pressures in terms of economic and environmental sustainability and also the increasing demands to attain the highest standards of health and welfare on farms.
Whilst some advances have been made in the management of common endemic diseases during the last 25 years, the incidence of specific endemic diseases remain unacceptably high. Key endemic conditions with substantive welfare and economic sequelae include lameness, mastitis, ketosis, metritis and these have an estimated annual prevalence in the UK of 37%, 32%, 30%, 10% respectively. The total cost of endemic disease to the UK dairy industry is estimated at £550M/annum (globally $80Bn/annum).
The transition and early lactation period (30 days pre-calving to 60 days post-calving) is a critical and demanding phase for dairy cows, during which cows undergo significant hormonal, metabolic, immunological, and physiological changes. Approximately 75% of disease risk (e.g., lameness, mastitis, metritis) is attributed to this period and approximately 50% of cows are impacted by a transition-related condition. There is wealth of evidence to suggest that these diseases are interrelated. Disease interactions are also important in defining productivity and reproductive outcomes for the cow. To create a step change in managing endemic disease on-farm, we believe farmers require holistic solutions that do not simply focus on individual diseases but that cover all endemic disease and include sustainable production. Such solutions will allow farmers to optimise breeding, culling, treatment and preventive decisions. No such holistic tools exist to date.
To date, research related to predicting cow health using transition period markers and technologies has generally shown models are mediocre. One key reason for this is that the methods employed have used basic, static features from data signals rather the dynamic signal properties and also have explored limited range of features. Next generation digital health platforms in humans are growing exponentially and their success is largely due to utilisation of varied range of dynamic time series data to evaluate complex signals on a longitudinal basis. These next generation features or ‘resilience indicators’ provide key information to predict health, longevity and well-being.
Building on previous research, working with industry partners and by utilising existing and evolving technologies to create novel features, we propose an individual cow ‘transition signature’ can be developed from a novel combination of digital, genetic and biological data that will allow the prediction of disease vulnerability as well as insights into underlying biological mechanisms of health, and ii) a key targeted intervention with NSAIDs after parturition will mitigate the issue of subacute inflammation at calving and in turn lead to an improvement in cow health.
This study provide a completely new method to define and predict a holistic “sustainable health index” in dairy cow which will be utilised as digital platform for on farm decision making. Using co-creation stakeholder will input in design and evaluation of this platform- we will house this solution in an existing commercial solution REMEDY as developed by our industry partner.
Technical Summary
To create a step change in managing endemic disease on-farm, we suggest farmers require holistic solutions that do not simply focus on individual diseases but that cover all endemic disease and include sustainable production. Transition period (30 day before calving to upto 60 days post calving) plays a substantial role in dictating a cow’s future health, productivity and reproductive performance. In this project we will co-develop a digital platform that utilises novel features captured in ‘transition period” from existing and evolving technologies on commercial dairy farms to predict a stakeholder informed sustainable health index.
In Obj 1 we Undertake a field study on 6 commercial dairy farms (n=1200 cows) to
collect and quantify predictor variables collected in transition period; static and dynamic features and individual cow profiles including behavioural, physiological (including metabolic) and genetic information captured via various tools and technologies.
In Obj 2, We co-create our key outcome variable, the cow’s sustainability health index. It will be computed from three domain areas (health, production and reproduction) using novel methodologies from economics and sustainability (Concordet approach) and thresholds for individual indicators will be informed from stakeholders’ elicitation.
In Obj 3 We undertake quantitative analyses using machine learning and dynamic modelling to optimise the prediction of a sustainable health index using the suite of predictor variables collected during the transition period (Obj 1). We also conduct a randomised controlled trial to evaluate the impact of NSAID on the sustainable health index as well as on individual outcomes related to health, reproduction and production (Obj part 1)
In Obj 4 we finally Integrate final predictive models within a commercial platform (REMEDY) for immediate farm implementation
Contact details:
Professor Jasmeet Kaler
Professor in Epidemiology and Precision Livestock Informatics, Faculty of Medicine & Health Sciences
University of Nottingham