Table 4. Summary risk-of-bias assessment by study design.
Qualitative summary of risk-of-bias domains across included studies by study design, informed by principles from RoB 2 and NOS. Domains assessed included selection bias, confounding, and outcome assessment.
RoB 2: Cochrane Risk of Bias 2, NOS: Newcastle-Ottawa Scale
| Study design | Selection bias | Confounding | Outcome assessment | Overall risk of bias |
| Randomised controlled trials | Low | Low | Low | Low |
| Post-hoc analyses of phase III trials | Low–moderate | Moderate | Low | Moderate |
| Prospective observational cohort studies | Moderate | Moderate–high | Low–moderate | Moderate |
| Retrospective cohort and registry studies | Moderate | High | Moderate | Moderate–high |
| Mechanistic imaging and biomarker studies | Moderate | Moderate | Low | Moderate |