Development of a tool for rapid assessment of the evidence of human observational epidemiological studies with focus on risk of bias in the context of public health and risk assessment

Abstract This report outlines the methodologies and outcome of the project developed under Framework Partnership Agreement (FPA) No GP/EFSA/AMU/2020/02 – Specific Agreement No. 05 entitled “Critical appraisal tools (CATs) for evaluation of the evidence from human observational epidemiological studies and further use in weight‐of‐evidence approaches” by Federal Institute for Risk Assessment (BfR) and EFSA. A tool for the rapid assessment of risk of bias (raRoB) in observational epidemiological human studies was developed. Even though, numerous tools for assessing methodological quality have been published, only few are applicable to observational studies. In the context of risk assessment, a tool was created to assess the risk of bias that is easy to use, even for those with only basic epidemiological knowledge. This tool is transparent, relatively quick to use and consists of five domains (Selection, Exposure, Outcome, Confounding, Follow‐up, Analysis, Selective reporting). The tool itself was tested in three rounds with experts and step by step improved. The accompanying documentation was also revised due to expert's comments. The raRoB tool allows a transparent documentation of the assessment of human observational epidemiological human studies with items in each domain and a summary of the scores and an indication for the worst ratio. Expert can additionally write their appraisal for their scores in their own words. This allows transparency and comparability.