Interventions to change antimicrobial use in livestock: A scoping review and an impact pathway analysis of what works, how, for whom and why.

Antimicrobial resistance (AMR) is a public health threat responsible for 700,000 deaths per year worldwide. Antimicrobial use (AMU) in livestock contributes to AMR in animal and public health. Therefore, it is essential to implement effective interventions towards better AMU in livestock. However, there is a lack of evidence to inform decision-makers of what works, how, for whom and why and how effective interventions can be adapted to different contexts. We conducted a scoping review and an impact pathway analysis to systematically map the research done in this area and to inform evidence-based and context-appropriate policies. We followed the PRISMA-ScR requirements and searched Web of Science, PubMed and Scopus databases to identify studies in English or French languages, in open access and published between 2000 and 2022. We selected thirty references addressing twenty-eight different interventions that were successful in changing AMU in livestock. We used an impact pathway logic model as an analytic framework to guide the technical aspects of the scoping review process and to identify the complex relationships between outputs, outcomes, impacts and contextual factors. A majority of interventions managed to improve AMU by changing herd and health management practices (ni=18). We identified intermediate outcomes including change in the veterinarian-farmer relationship (ni=7), in knowledge and perception (ni=6), and in motivation and confidence (ni=1). Twenty-two studies recorded positive impacts on animal health and welfare (ni=11), technical performances (ni=9), economic performances (ni=4) and AMR reduction (ni=4). Interventions implemented different strategies including herd and health management support (ni=20), norms and standards (ni=11), informational and educational measures (ni=10), economic support (ni=5). Studies were mainly in European countries and in pig and large ruminants farming. Most interventions targeted farmers or veterinarians but we identified other major and influential actors including authority and governmental organizations, academics and research, organization of producers or veterinarians, herd advisors and technicians, laboratories, and public opinion. Key success factors were knowledge and perception (ni=14), social factors (ni=13), intervention characteristics (ni=11), trajectory and ecosystem of change (ni=11), economic factors (ni=9), herd and health status (ni=8), data access and monitoring (ni=4). This review describes a paucity of impact assessment of interventions towards better AMU in livestock. There is no one-size-fits-all transition pathway but we inform decision-makers about the most successful interventions that work, how, for whom and why. The impact pathway analysis provided a holistic view of the successful change processes and the complex relationships between outputs, outcomes, impacts and contexts.

Ranking of simultaneously presented choice options in animal preference experiments.

We considered experiments where animals chose one of all possible simultaneously presented options. The animals might be observed at repeated occasions. In the ethological literature the analysis is often focused on testing the global hypothesis of no difference in preferences by non-parametric methods. This fails to address the estimation of a ranking. Often this approach cannot adequately reflect the experimental setting and the repeated measurement structure. Therefore, we propose to model the choice probabilities for the options with a multinomial logistic model. The correlation induced by repeated measurements is incorporated by animal specific random intercepts. The ranking of the options is taken as the order of the choice probabilities. Adopting a Bayesian approach samples from the posterior distribution of the choice probabilities provide directly samples from the posterior of the rankings. Based on this an estimate of the ranking and description of its variability can be derived. The computation was performed via Markov chain Monte Carlo sampling and was implemented using WinBUGS. We illustrate our approach with an experiment to determine the preference of pigs for three different rooting materials. The proposed method allowed deriving an overall ranking for different combinations of the materials and the spatial positioning.