Ten tips for successfully assessing risk of bias in randomised control trials

Users of RoB 2 – Cochrane’s risk of bias assessment tool for randomised control trials – frequently use it in ways developers hadn’t intended, despite detailed guidance, webinars and FAQs being available. This can lead to overly harsh or lenient assessments of bias and can change how confident researchers are about the evidence base of randomised trials.

The authors of a paper published in Cochrane Evidence Synthesis and Methods worked as systematic review editors for Cochrane for 19 months. Based on their observations of RoB 2 misuse, they compiled a list of top 10 tips for successfully assessing risk of bias using the tool.

Suggestions are broken down into three helpful categories: planning, applying the tool, and resolving common problems and are presented in the paper with links to detailed guidance.

The paper also highlights useful training resources, including webinars, and advises author teams to encourage one member to become a RoB 2 expert or seek support from a more experienced user.

Theresa Moore, Senior Research Associate at NIHR ARC West and lead author, said:

“In this paper we highlighted the 10 main issues we observed when colleagues used RoB 2.

“We did this with the aim of optimising how RoB 2 is used, avoiding some of the frequent misapplications of the tool and demonstrating how to present RoB 2 judgements within a review.

“Overall, we would encourage users to access training resources, including webinars.

“Although the information presented in our paper was drawn from Cochrane reviews, the issues we raise and our suggestions for improving how RoB 2 is used are likely to be relevant to authors of all systematic reviews that include randomised trials and use the tool.”