Severity assessment in animal models for colitis and surgery – validation, translation and refinement

• Applicant: Professor Dr. André Bleich
• Subject Area: General and Visceral Surgery ; Sensory and Behavioural Biology ; Veterinary Medical Science ; Animal Breeding, Animal Nutrition, Animal Husbandry
• Term: since 2017
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 321137804

In the first two funding periods, our objective was to evaluate objective and advanced methods for the assessment of severity in colitis and surgical models. We were able to identify reliable variables such as body weight, voluntary wheel running, heart rate, burrowing behaviour, and the Mouse Grimace Scale (MGS) and validated these parameters across diverse models. Together with the data science project of FOR 2591, we established algorithm-based multimodal severity assessment; e.g., we developed together a composite metric called the RELative Severity Assessment (RELSA), which enabled the comparison across different model systems, utilising clinical, physiological, and behavioural data. Using these approaches, we were able to effectively assess severity in various model systems and compare directly in a quantitative manner. Using these approaches, we assessed and implemented refinement approaches, including the evaluation of pain management, analgesic treatments, and non-pain-related measures. We developed group-housing test systems and explored handling methods to minimise the overall burden during experimentation. In the upcoming funding period, we plan to further analyse the core parameters for severity assessment and their impact on refining experimental procedures across FOR 2591. This will involve evaluating the effectiveness of validated computational approaches, refining cage-side scoring through algorithm-based analysis, and assessing the impact of refinement measures such as model-specific analgesia. Our main goals are to evaluate the replicability, robustness, and generalisability of the variables and review refinement options in experiments and humane endpoints. With other consortium members, we aim to develop a Severity Tool Box that encompasses prospective, current, and retrospective severity assessment in order to provide evidence-based recommendations for refinement options and humane endpoints. The project will address the concept of mutivariate severity assessment, method validation, and the further development of current technology for collecting behavioural and physiological data, to provide science and authorities with comprehensive validated methods that are increasingly demanded. Focused group discussions will also enable us to align procedures and models assessed using advanced parameters and algorithms with existing severity classes. The project plans to suggest boundaries through data science methods and ultimately define them through inter-group agreement using a Delphi process. The findings will be summarised and reference data developed, establishing the recommendations for the community. Together with the consortium, these recommendations will be disseminated locally, nationally, internationally through traditional and novel forms of public outreach.

• DFG Programme: Research Units
• Subproject of: FOR 2591: Severity assessment in animal-based research