MAPRA: Modelling animal pathogens. Review and adaptation.

Epidemiological models are frequently used by EFSA as a support tool to quantify the risk of introduction, establishment, spread and transmission of diseases of importance to livestock and wildlife in the EU. Modelling approaches from fields outside veterinary epidemiology might also be applicable, and may provide new ideas that could be leveraged to improve the utility of models used by EFSA. This project was set up to identify models that are currently used within all relevant fields, and to build an inventory describing their characteristics. A series of expert workshops were held to obtain insights into unfamiliar research fields, and to build a dictionary of terms for a literature search within each identified field. A subsequent search of the literature identified 3,625 potentially applicable papers matching the search criteria, of which 468 passed a second screening phase and went on to be extracted into the final inventory. A secondary aim of the project was to demonstrate the potential use of this inventory in developing new models to be applied to three case studies: the introduction of peste des petits ruminants (PPR), the establishment of African swine fever (ASF), and the spread of bluetongue (BT). The novel PPR model was developed to propagate uncertainty associated with a data-sparse environment, and incorporates ideas from graph theory and operations research to obtain an efficient representation of disease spread. Two novel models were developed for ASF, based on spatial K-functions and using INLA to model the complex spatial structure in the absence of denominator data. A novel modelling framework was developed for bluetongue to implement the highly complex mechanistic model incorporating a high level of data availability with a flexible network representation of between-farm disease spread. For each case study, the novel models compared favourably to the existing models that represented the previous state-of-the-art modelling approaches