A new instrument to assess Risk of Bias in Non-randomised Studies of Exposures (ROBINS-E): Application to studies of environmental exposure

Article type
Authors
Morgan R1, Sterne J2, Higgins J3, Thayer K4, Schunemann H1, Rooney A5, Taylor K5
1McMaster University
2University of Bristol, UK
3University of Bristol
4Environmental Protection Agency
5National Institute for Environmental Health Sciences
Abstract
Objectives: To introduce a new tool for assessing risk of bias in non-randomised studies of exposures, and illustrate its application to studies of environmental and occupational exposure.

Description: Systematic reviews should include rigorous risk-of-bias assessments of included studies. We are adapting the recently published ROBINS-I instrument (for risk of bias in non-randomised studies of interventions) to address non-randomized studies of exposures other than interventions, including environmental and occupational exposures. This workshop will present an overview of the adapted instrument (to be named ROBINS-E), and will explain the proposed changes to ROBINS-I that will make ROBINS-E more suitable for assessing studies of exposures). Key aspects include preliminary consideration of risk of bias within the review protocol, use of signaling questions to inform risk-of-bias judgments, specification of a 'target experiment', detailed assessments of confounding and exposure measurement error, and guidance for interpretation across a body of evidence.
There will be a hands-on exercise to apply the draft ROBINS-E instrument to individual studies of a specific environmental exposure. The workshop will conclude by making a study-level risk-of-bias judgment, and discuss considerations when using ROBINS-E to assess individual studies to inform a systematic review, including making an overall risk-of-bias judgment across a body of evidence for a specific outcome.