Critical Appraisal Workspace

Risk of bias, quality assessment & GRADE — in one place

← Back to home

RoB 2 (randomised trials)

Study label Overall

Add study …

Traffic-light plot

Summary plot

ROBINS-I (non-randomised studies)

Study label Overall

Add study …

Traffic-light plot

Summary plot

Newcastle–Ottawa Scale

Score each study item by item; domain stars, the total and the quality band update as you answer.

Add study …

Studies assessed
Mean stars
Median stars

Star table

One column per criterion, with domain subtotals and the total out of the scale maximum — the table journals print.

Summary plot

GRADE

Outcome Study design Certainty of evidence

Add outcome …

Upgrading factors

Observational evidence starts at Low. These three factors can raise the certainty again. A randomised trial already starts at High, so they do not apply to one.

GRADE Summary of Findings

The certainty columns come from the assessment above. Type the event counts and the pooled relative effect; the absolute risks are worked out from them.

Outcome Certainty assessment Summary of findings
Participants (studies) Other considerations Certainty of evidence Relative effect (95% CI)

GRADE certainty: ⨁⨁⨁⨁ High · ⨁⨁⨁◯ Moderate · ⨁⨁◯◯ Low · ⨁◯◯◯ Very low. CI: confidence interval.

The risk with the comparator is the observed control-arm event rate unless you state an assumed risk. The risk difference is derived from that risk and the relative effect — an odds ratio is converted through the odds, not applied to the risk directly.