Simin Lee ^1* Christopher Baker ^2,3,4 Emily Sellens ⁵ Mark Stevenson ¹ Sharon Roche ⁵ Robyn N. Hall ⁶ Andrew C. Breed ⁵ Simon Firestone ¹

• ¹ Faculty of Veterinary and Agricultural Sciences, The University of Melbourne, Parkville, Victoria, Australia
• ² School of Mathematics and Statistics, Faculty of Science, University of Melbourne, Parkville, Victoria, Australia
• ³ Melbourne Centre for Data Science, Faculty of Science, University of Melbourne, Melbourne, Victoria, Australia
• ⁴ Centre of Excellence for Biosecurity Risk Analysis, School of Biosciences, Faculty of Science, University of Melbourne, Melbourne, Victoria, Australia
• ⁵ Department of Agriculture, Fisheries and Forestry, Canberra, Australian Capital Territory, Australia
• ⁶ Ausvet, Canberra, Australian Capital Territory, Australia

Lumpy skin disease (LSD) is an infectious disease currently spreading worldwide and poses a serious global threat. However, there is limited evidence and understanding to support the use of models to inform decision-making in LSD outbreak responses. This review aimed to identify modelling approaches that can be used before and during an outbreak of LSD, examining their characteristics and priorities, and proposing a structured workflow. We conducted a systematic review and identified 51 relevant publications on LSD outbreak modelling. The review identified six categories of question to be addressed following outbreak detection (origin, entry pathway, outbreak severity, risk factors, spread, and effectiveness of control measures), and five analytical techniques used to address them (descriptive epidemiology, risk factor analysis, spatiotemporal analysis, dynamic transmission modelling, and simulation modelling). We evaluated the questions each analytical technique can address, along with their data requirements and limitations, and accordingly assigned priorities to the modelling. Based on this, we propose a structured workflow for modelling during an LSD outbreak. Additionally, we emphasise the importance of pre-outbreak preparation and continuous updating of modelling post-outbreak for effective decision-making. This study also discusses the inherent limitations and uncertainties in the identified modelling approaches. To support this workflow, highquality data must be collected in standardised formats, and efforts should be made to reduce inherent uncertainties of the models. The suggested modelling workflow can be used as a process to support rapid response for countries facing their first LSD occurrence and can be adapted to other transboundary diseases.