RigorMed
← Research resources

Research methods guide

How to choose RoB 2 or ROBINS-I

Match the tool to the design and result. Consider the relevant RoB 2 variant for randomized trials, ROBINS-I for nonrandomized intervention studies, and a suitable tool such as ROBINS-E for nonrandomized exposure studies.

Reviews and evidence synthesis

A review includes randomized trials and observational comparisons of interventions. Choose a tool by design, target effect and specific result.

Choosing a risk-of-bias tool by design

Record the instrument version and module in the protocol.

Choosing a risk-of-bias tool by design
ItemWorked exampleVerification
Parallel randomized trial12-week BP change, effect of assignment. Use the appropriate RoB 2 version.Inspect randomization, missingness, measurement and selective reporting with source locations.
Nonrandomized intervention comparisonCompare participation in an intervention where severity affects assignment. Consider the applicable ROBINS-I version.Define target trial, time zero and key confounders.
Diagnostic accuracyEvaluate sensitivity and specificity of a screening test.Check the appropriate QUADAS version and protocol for index test, reference standard and flow.

Apply the method to your materials

  1. Assess the specific estimate

    Key by study, outcome, time, comparison and target effect. Outcomes in one paper can have different bias judgments.

  2. Save signaling answers and rationale

    Assess independently using the same version. Record source text, adjudication and synthesis implications.

Questions that arise in practice

Can the tools be converted to one score?

Report domains and judgments on their own terms, then explain implications by design and conclusion.

Verified sources

The official sources below were checked on 2026-11-12. Access conditions and update dates are rechecked before publication.

  1. Risk of Bias tools
  2. Cochrane Handbook, Chapter 3

RigorMed workspace

Start with your research materials

Bring your research question, available materials, and the steps you want to check.