0 avis
Combining Dynamic Generalized Linear Models And Mechanistic Modelling To Optimize Treatment Strategies For Bovine Respiratory Disease In Cattle Fattening Farms
Archive ouverte
Edité par CCSD -
International audience. Bovine Respiratory Disease (BRD) is a major concern for young bulls’ health and welfare. The use of metaphylaxis, a treatment strategy that aims to prevent BRD outbreaks by treating the entire batch (collective treatment), has been widely adopted to minimize economic losses. However, the decision to perform collective treatments involves a trade-off between the cumulative incidence of BRD cases and antimicrobial usage (AMU), which raises concerns about the optimal timing of treatment. To address this challenge, we developed a proof of concept of a decision support tool to assist farmers and veterinarians in making informed decisions about when to perform collective treatment for BRD. The proposed tool consists of a framework that combines a mechanistic stochastic simulation engine modeling the spread of a BRD pathogen (Mannheimia haemolytica) and a dynamic generalized linear model (DGLM) that provides early warnings based on observed data such as clinical signs, here using synthetic data. We looked at 24 scenarios in total, combining four individual risk levels of BRD, two types of allocation systems in batches (sorted by risk level or randomly allocated), and three interventions (individual treatments, conventional collective treatments, and collective treatments triggered by the DGLM early warnings). Our results demonstrated that collective treatments triggered by early warnings were the most effective strategy for reducing the cumulative incidence of BRD cases, particularly in high-risk herds. Individual treatments had the smallest AMU but had the biggest median cumulative incidence in all combinations of risk levels and types of animal allocation in batches. Collective treatments triggered by early warnings had lower AMU in all scenarios when compared to conventional collective treatments. Our findings suggest that the proposed decision support tool can provide valuable guidance in reducing BRD cases and AMU. Future work will focus on refining the tool to incorporate real-world data.
Consulter en ligne
Suggestions
Du même auteur
