Critical appraisal tool (CAT) for the evaluation of evidence from animal observational epidemiological studies

Abstract

Risk of bias assessments in animal observational epidemiological studies can be challenging due to the heterogeneity in observational study designs and the lack of available tools for guided approaches. We modified the structure of the previously developed tool for observational human studies, making it applicable to animal observational epidemiological studies and allowing researchers to conduct a systematic, comprehensive, standardized evaluation of the risk of bias in animal observational epidemiological studies. The prototype tool development was following a three-stage consensus-building approach: 1) defining the scope and overall structure of the tool. 2) articulating the wording of domains and items and their signalling questions. 3) finalising the tool with a test round to ensure its reliability and user-friendliness. Each stage consisted of a Delphi-round followed by an online consensus group meeting in a panel of experts. We engaged a total of 13 experts from different areas of veterinary epidemiology, tool development, and reproducible research methodology. The modified Delphi study allowed the consensus-based development of the tool. Seven experts tested the prototype in a pilot study. The tool was then programmed as a web app and its usability was determined. Results from the test phase were then implemented into a final version of the prototype tool. The raRoB-vet tool was developed for evaluation of animal observational studies to allow for standardized and reproducible assessment of risk of bias. It contributes to fill the existing gap in critical appraisal tools in epidemiological animal studies. The tool is currently considered a prototype tool that will benefit from further testing of its feasibility, validity, reliability, reproducibility and fitness for specific purposes.