Cited by 128Meta-analysis and systematic reviewsAdvanced Causal Inference TechniquesStatistical Methods in Epidemiology

The DOI matches its Crossref record; 23 of the 23 references it cites were found in a registry or an index; 1 of the 14 sources its claims cite were checked for support.

RoBANS 2: A Revised Risk of Bias Assessment Tool for Nonrandomized Studies of Interventions

Open access
Seo, Hyun-Ju, Kim, Soo Young, Lee, Yoon Jae, Park, Ji-Eun·2023·Korean Journal of Family Medicine

Abstract

Assessment of the risk of bias is an essential component of any systematic review. This is true for both nonrandomized studies and randomized trials, which are the main study designs of systematic reviews. The Risk of Bias Assessment Tool for Nonrandomized Studies (RoBANS) was developed in 2013 and has gained wide usage as a risk-of-bias assessment tool for nonrandomized studies. Four risk-of-bias assessment experts revised it by reviewing existing assessment tools and user surveys. The main modifications included additional domains of selection and detection bias susceptible to nonrandomized studies of interventions, a more detailed consideration of the comparability of participants, and more reliable and valid outcome measurements. A psychometric assessment of the revised RoBANS (RoBANS 2) revealed acceptable inter-rater reliability (weighted kappa, 0.25 to 0.49) and construct validity in which intervention effects of studies with an unclear or high risk of bias were overestimated. The RoBANS 2 has acceptable feasibility, fair-to-moderate reliability, and construct validity. It provides a comprehensive framework for allowing authors to assess and understand the plausible risk of bias in nonrandomized studies of interventions.

Overview

AI snapshot

Written from the paper’s full text.

TL;DR

RoBANS 2 is a revised outcome-based risk-of-bias tool for nonrandomized studies of interventions, adding selection and detection bias domains and more detailed consideration of participant comparability and outcome measurement. In psychometric testing it showed acceptable feasibility, fair-to-moderate inter-rater reliability (weighted kappa 0.25–0.49), and construct validity, with unclear/high risk-of-bias studies overestimating intervention effects.

Key findings

  • RoBANS 2 revised the original RoBANS by adding domains for selection and detection bias, separating participant selection into comparability of participants and target group selection, and revising outcome assessment to consider blinding of outcome assessors and reliable/valid outcome measurement methods.
  • Feasibility: mean time to complete risk-of-bias assessment was 8.72±5.00 minutes per article, with a range from 20 seconds to 36.35 minutes.
  • Inter-rater reliability: all eight RoBANS 2 domains had fair agreement or higher, with weighted kappa statistics ranging from 0.25 to 0.49.
  • Construct validity: among 112 nonrandomized studies included for inter-rater reliability, 77 were included for construct validity; studies conducted inadequately for each risk-of-bias domain were likely to report low odds ratios in seven of eight domains, indicating that intervention effects with unclear or high risk of bias were overestimated.

Methods

  • The revision reviewed existing risk-of-bias/critical appraisal checklists (SIGN, Newcastle-Ottawa Scale, AHRQ checklist, RTI Item Bank) and held consultative meetings with five systematic review methodologists; a sample of nonrandomized studies was compiled via the National Evidence-based Healthcare Collaborating Agency, Health Insurance Review and Assessment Service, PubMed, and the Cochrane Library.
  • Inclusion criteria for psychometric evaluation were studies with no intervention or placebo control, dichotomous outcome data except before-and-after studies, and designs including cohort, case-control, cross-sectional, or before-and-after studies; 112 studies were selected: 45 cohort, 16 case-control, 25 cross-sectional, and 26 before-and-after. The minimum required by two raters was 85 studies, based on 80% power to detect a kappa of 0.70 against a null kappa of 0.40.
  • Paired assessors with doctoral degrees and at least 10 years of systematic review experience independently rated studies after pilot testing; analyses used weighted kappa for reliability and compared effect sizes by risk-of-bias domain using Cohen’s d for continuous outcomes and Hasselblad and Hedges’s transformation for odds ratios, with SAS 9.4, Stata SE 16.0, and RevMan 5.4; tests were two-sided at α=0.05.

Limitations

  • Stated in the paper: further refinement and extensive feedback from RoBANS 2 users are required; for controlled before-after studies, interrupted time series, and interrupted time series with comparisons, assessors need to determine domain judgments considering study design nature from epidemiological experts; the authors recommend Cochrane risk-of-bias tools for RCTs, non-RCTs, or quasi-experimental trials; further research is needed to compare inter-rater agreement and usability of RoBANS 2 and ROBINS-I for cohort-type designs.
  • Evident from the text: reliability was only fair-to-moderate (weighted kappa 0.25–0.49), not substantial or almost perfect; construct validity analysis excluded 26 before-and-after studies without comparison and nine studies unable to extract data, leaving 77 of 112 studies; no direct comparison with ROBINS-I was performed in this study.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.4082/kjfm.23.0034

Cites

23 of 23 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Verification

Claims the paper makes

1 of 14 cited sources checked · 0 supported · 0 flagged · accuracy not yet measured

  • Findingown data

    All eight domains of RoBANS 2 had fair agreement or higher.

    “All domains of eight RoBANS 2 had fair agreement or higher, ranging from 0.25 to 0.49 κ statistics.”
    Quote in the paperSubstantial claim
  • Findingown data

    The inter-rater reliability of RoBANS 2 ranged from 0.25 to 0.49 weighted kappa.

    “All domains of eight RoBANS 2 had fair agreement or higher, ranging from 0.25 to 0.49 κ statistics.”
    Quote in the paperSubstantial claim
  • Findingown data

    Studies conducted inadequately for each domain of the risk of bias were likely to report low ORs in seven of eight domains.

    “Our findings revealed that studies conducted inadequately for each domain of the risk of bias were likely to report low ORs in seven of eight domains (Table 3).”
    Quote in the paperSubstantial claim
  • Findingnone given

    The tools RoBANS 2 and ROBINS-I overlap substantially in terms of the risk of bias domains.

    “However, the tools overlap substantially in terms of the risk of bias domains (Appendix 3).”
    Quote in the paperNot a substantial claim
  • Study design and conductcitation

    The minimum number of studies required by the two raters was 85, based on a dichotomous variable with 80% power, to detect a kappa of 0.70, at a proportion of positive ratings of 0.70.

    “The minimum number of studies required by the two raters was 85, based on a dichotomous variable with 80% power, to detect a kappa of 0.70, at a proportion of positive ratings of 0.70.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [12] source
  • It may state two separate claims as one. 9
  • Probably not a substantial claim. 9
  • It may state two separate claims as one.

    Premisecitation

    A crucial limitation of observational studies on intervention effects and adverse reactions is that the intervention of interest is not randomly assigned, blinding is lacking, and there is often no comparison.

    “However, a crucial limitation of observational studies on intervention effects and adverse reactions is that the intervention of interest is not randomly assigned, blinding is lacking, and there is often no comparison.”
    Quote in the paperSubstantial claim1 of 2 sources checked
    • cites [5] source
    • cites [6] source

      The 'Inclusion of non-randomised studies in systematic reviews' section says non-randomised intervention studies face biases that can make their effect estimates inaccurate, supporting the concern about non-random assignment. It does not say that observational studies generally lack blinding or often have no comparison group. (LLM, conf=0.86)

      Experimental — accuracy not yet measured
  • It may state two separate claims as one.

    Findingown data

    The mean time to complete the risk of bias assessment using RoBANS 2 was 8.72±5.00 minutes per article, with a range of 20 seconds to 36.35 minutes.

    “The time spent assessing each study ranged from 20 seconds to 36.35 minutes, with a mean of 8.72±5.00 minutes per article.”
    Quote in the paperSubstantial claim
  • It may state two separate claims as one.

    Findingnone given

    The main modifications in RoBANS 2 included additional domains of selection and detection bias susceptible to nonrandomized studies of interventions, a more detailed consideration of the comparability of participants, and more reliable and valid outcome measurements.

    “The main modifications included additional domains of selection and detection bias susceptible to nonrandomized studies of interventions, a more detailed consideration of the comparability of participants, and more reliable and valid outcome measurements.”
    • Also stated in: § DISCUSSION · plain
      “The main modifications included additional domains of selection and detection bias susceptible to the NRSI, a more detailed consideration of the comparability of participants, and more reliable and valid outcome measurements.”
    Quote in the paperSubstantial claim
  • Probably not a substantial claim.

    Findingnone given

    Similar to the original RoBANS, the assessments in RoBANS 2 were related to the risk of bias in the estimates of the intervention effect for a single outcome or endpoint rather than at the study level.

    “Similar to the original RoBANS, the assessments in RoBANS 2 were related to the risk of bias in the estimates of the intervention effect for a single outcome or endpoint rather than at the study level.”
    Quote in the paperSubstantial claim
  • Probably not a substantial claim. The paper’s conclusions may not depend on it.

    Findingnone given

    Similar to the original version, RoBANS 2 is an outcome-based checklist.

    “Similar to the original version, RoBANS 2 is an outcome-based checklist.”
    Quote in the paperNot a substantial claim
  • Probably not a substantial claim.

    Findingreasoning

    The domains of blinding of outcome assessors, outcome assessment, and incomplete outcome data can be treated as result-based evaluations because they are classified as patient-reported outcomes or objective outcome measures.

    “Additionally, the domains of blinding of outcome assessors, outcome assessment, and incomplete outcome data can be treated as result-based evaluations because they are classified as patient-reported outcomes or objective outcome measures.”
    Quote in the paperSubstantial claim
  • It may state two separate claims as one.

    Interpretationreasoning

    RoBANS 2 has construct validity and can detect significant differences in effect size estimates according to the risk of bias.

    “Therefore, the RoBANS 2 has construct validity and can detect significant differences in effect size estimates according to the risk of bias.”
    Quote in the paperSubstantial claim
  • Probably not a substantial claim.

    Study design and conductown data

    Paired assessors independently evaluated the risk of bias of the included studies using RoBANS 2 after pilot testing a sample of included studies.

    “Paired assessors of the review team independently evaluated the risk of bias of the included studies using RoBANS 2 after pilot testing a sample of included studies.”
    Quote in the paperSubstantial claim
  • It may state two separate claims as one.

    Premisecitation

    RoBANS was developed in 2013 and has gained wide usage as a risk-of-bias assessment tool for nonrandomized studies.

    “The Risk of Bias Assessment Tool for Nonrandomized Studies (RoBANS) was developed in 2013 and has gained wide usage as a risk-of-bias assessment tool for nonrandomized studies.”
    • Also stated in: § INTRODUCTION · plain
      “The Risk of Bias Assessment Tool for Nonrandomized Studies (RoBANS), published in 2013 [7], is a widely used bias tool.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [7] source
  • Probably not a substantial claim.

    Premisecitation

    In nonrandomized studies, selection bias occurs when participants chosen for the intervention of interest have different characteristics from those allocated to the alternative intervention (or not treated) because the choice of a given intervention might be affected by the discretion of the treating clinician or patient preference, patient characteristics, and clinical history.

    “In nonrandomized studies, selection bias occurs when participants chosen for the intervention of interest have different characteristics from those allocated to the alternative intervention (or not treated) because the choice of a given intervention might be affected by the discretion of the treating clinician or patient preference, patient characteristics, and clinical history [13].”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [13] source
  • Probably not a substantial claim. The paper’s conclusions may not depend on it.

    Study design and conductnone given

    Each study was randomly assigned to paired assessors using computer-based random number generation.

    “Each study was randomly assigned to paired assessors using computer-based random number generation.”
    Quote in the paperNot a substantial claim
  • It may state two separate claims as one.

    Interpretationown data

    RoBANS 2 had acceptable feasibility, fair to moderate reliability, and construct validity.

    “In conclusion, RoBANS 2 had acceptable feasibility, fair to moderate reliability, and construct validity.”
    • Also stated in: § Abstract · plain
      “The RoBANS 2 has acceptable feasibility, fair-to-moderate reliability, and construct validity.”
    Quote in the paperSubstantial claim
  • Probably not a substantial claim.

    Premisecitation

    Real-world evidence has become increasingly critical for identifying the effects and safety of healthcare interventions.

    “Real-world evidence has become increasingly critical for identifying the effects and safety of healthcare interventions.”
    Quote in the paperSubstantial claim0 of 1 sources checked
  • It may state two separate claims as one.

    Premisecitation

    The findings of nonrandomized studies of interventions can be generalized to the population and provide clinical evidence regarding the benefits or risks of healthcare interventions.

    “The findings of nonrandomized studies of interventions (NRSI) can be generalized to the population and provide clinical evidence regarding the benefits or risks of healthcare interventions [1].”
    Quote in the paperSubstantial claim0 of 1 sources checked
  • Probably not a substantial claim. The paper’s conclusions may not depend on it.

    Premisecitation

    These reviews may allow for the evaluation of adverse events and long-term effects after exposure to healthcare interventions.

    “Moreover, these reviews may allow for the evaluation of adverse events and long-term effects after exposure to healthcare interventions [3].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
  • It may state two separate claims as one.

    Premisecitation

    The study findings of observational studies are susceptible to confounding and selection biases, which could result in biased estimates of intervention effects compared with smaller RCTs.

    “Thus, the study findings are susceptible to confounding and selection biases, which could result in biased estimates of intervention effects compared with smaller RCTs [5,6].”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [5] source
    • cites [6] source
  • Probably not a substantial claim. The paper’s conclusions may not depend on it.

    Premisecitation

    Researchers have increasingly included NRSI in systematic reviews to examine various interventions such as medications, hospital procedures, community health interventions, and health systems.

    “Researchers have increasingly included NRSI in systematic reviews to examine various interventions such as medications, hospital procedures, community health interventions, and health systems [2].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [2]

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • It may state two separate claims as one.

    Premisecitation

    Bias can occur when outcome assessors are aware of the intervention status, if different methods are used to assess outcomes in different intervention groups, or if measurement errors are related to the intervention status or effects.

    “Bias can occur when outcome assessors are aware of the intervention status, if different methods are used to assess outcomes in different intervention groups, or if measurement errors are related to the intervention status or effects [16].”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [16] source
18 of 18

Citations

Cites

Flagged first, then by citations

Research works printed but not linked: 4

  1. c2023, Checklists [Internet]
  2. Wells, c2021, The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses [Internet]
  3. Warrens, 2013, Weighted kappas for 3×3 tables, 325831
  4. Review Manager (RevMan): version 5.4 [Internet]

Cited by

Sorted by citations

Metadata

Authors
Seo, H.-J., Kim, S. Y., Lee, Y. J., & Park, J.-E.
Affiliations
  • Chungnam National University (KR)
  • Hallym University (KR)
  • Jaseng Medical Foundation (KR)
Corresponding author
Kim, Soo Young
Venue
Korean Journal of Family Medicine
Published
2023-09-20
Volume
44
Issue
5
Pages
249-260
ISSN
  • 2005-6443 (ISSN-L)
  • 2092-6715
OpenAlex
W4383550757
Cited by
128
References
23
Citations by year
2026: 55 · 2025: 52 · 2024: 21
FWCI
13.06
Type
journal article
Language
en
Open access
Locations
  • Korean Journal of Family Medicine (publishedVersion, cc-by-nc)
  • PubMed Central (publishedVersion, cc-by-nc)
  • PubMed (publishedVersion)
Funders
Health Insurance Review and Assessment Service
Topics
  • Social Sciences › Decision Sciences › Statistics, Probability and Uncertainty › Meta-analysis and systematic reviews
  • Physical Sciences › Mathematics › Statistics and Probability › Advanced Causal Inference Techniques
  • Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods in Epidemiology
Keywords
risk of bias assessment, non-randomized studies, selection bias, detection bias, outcome measure validity, inter-rater reliability, construct validity, psychometric properties, systematic review methodology

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

A language model drew the claims from the paper on 2026-11-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-11-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review