ROBINS-I
ROBINS-I is a tool designed to assess how reliable non-randomised studies are when evaluating the effects of healthcare interventions. It helps identify potential biases that could affect the study's results.
At a glance
Use when
Conducting systematic reviews that include non-randomised studies of interventions, especially when assessing comparative effectiveness or safety.
Avoid when
Assessing randomised controlled trials or when insufficient study detail is available to make informed judgments about bias.
Inputs
Non-randomised study evaluating an intervention, including details on study design, population, intervention, comparators, outcomes, and analysis methods.
Outputs
Structured risk of bias assessment across seven domains, with an overall risk of bias judgment for the study.
How it works
ROBINS-I (Risk Of Bias In Non-randomised Studies - of Interventions) is a structured tool for assessing risk of bias in non-randomised studies estimating the comparative effectiveness of interventions. It evaluates bias across seven domains: confounding, selection of participants, classification of interventions, deviations from intended interventions, missing data, measurement of outcomes, and selection of reported results. Each domain is judged for risk of bias, leading to an overall risk of bias rating for the study. It supports systematic reviewers in critically appraising non-randomised studies included in evidence syntheses.
- HTA domains
- Clinical Effectiveness
- Categories
- Data Quality Assessment
- Assumptions
- The tool assumes that systematic differences between groups in non-randomised studies can lead to biased effect estimates, and that these biases can be systematically identified and evaluated.
- Strengths
- Provides a comprehensive, structured framework for bias assessment in non-randomised studies; supports transparency and consistency in systematic reviews; developed through expert consensus and empirical evidence.
- Limitations
- Requires substantial expertise to apply correctly; may be time-consuming; subjective judgments are involved in assessing bias domains.
- Also known as
- Risk Of Bias In Non-randomised Studies - of Interventions, ROBINS-I tool
Questions this answers
- › Is the study's estimate of the intervention effect likely to be biased due to confounding?
- › Were participants selected in a way that could introduce bias?
- › How accurately were interventions classified?
- › Did deviations from the intended interventions affect the results?
- › Could missing data have influenced the findings?
- › Was the outcome measured in a way that could bias the results?
References & sources
Related methods
Beta record. Generated from the primary source via AI extraction and independent audit, pending final human review.