article Open AccessTop 1% cited

ROBINS-I: a tool for assessing risk of bias in non-randomised studies of interventions

BMJ · 2016 · Vol. 355 · pp. i4919–i4919

Jonathan A C Sterne✉(University of Bristol)Miguel A. Hernán(Harvard–MIT Division of Health Sciences and Technology)Barnaby C Reeves(University of Bristol)Jelena Savović(University Hospitals Bristol NHS Foundation Trust)Nancy D Berkman(RTI International)Meera Viswanathan(RTI International)David Henry(Public Health Ontario)Douglas G. Altman(Nuffield Orthopaedic Centre)Mohammed Ansari(University of Ottawa)Isabelle Boutron(Sorbonne Paris Cité)James R. Carpenter(London School of Hygiene & Tropical Medicine)An‐Wen Chan(University of Toronto)Rachel Churchill(University of York)Jonathan J Deeks(University of Birmingham)Asbjørn Hróbjartsson(University of Southern Denmark)Jamie J Kirkham(University of Liverpool)Peter Jüni(University of Toronto)Yoon K. Loke(University of East Anglia)Theresa D Pigott(Loyola University Chicago)Craig Ramsay(University of Aberdeen)Deborah L. Regidor(Kaiser Permanente)Hannah R. Rothstein(Baruch College)Lakhbir Sandhu(University of Toronto)Pasqualina Santaguida(McMaster University)Holger J. Schünemann(Cochrane)Beverly Shea(University of Ottawa)Ian Shrier(McGill University)Peter Tugwell(University of Ottawa)Lucy Turner(Ottawa Hospital Research Institute)Jeffrey C. Valentine(University of Louisville)Hugh Waddington(London International Development Centre)Elizabeth Waters(University of Melbourne)George A. Wells(University of Ottawa)Penny Whiting(University of Bristol)Julian P. T. Higgins(University of Bristol)

View via DOI Full-text PDF

Abstract

Non-randomised studies of the effects of interventions are critical to many areas of healthcare evaluation, but their results may be biased. It is therefore important to understand and appraise their strengths and weaknesses. We developed ROBINS-I (“Risk Of Bias In Non-randomised Studies - of Interventions”), a new tool for evaluating risk of bias in estimates of the comparative effectiveness (harm or benefit) of interventions from studies that did not use randomisation to allocate units (individuals or clusters of individuals) to comparison groups. The tool will be particularly useful to those undertaking systematic reviews that include non-randomised studies.

Health Systems, Economic Evaluations, Quality of Life Meta-analysis and systematic reviews Primary Care and Health OutcomesPsychological interventionMedicineStrengths and weaknessesSystematic reviewHarmMeta-analysisMEDLINEPsychologyNursingPathology