Table 1

Criteria for downgrading the quality of outcomes using the GRADE approach

Reason to downgrade the level of evidence
Risk of bias

  • Majority of studies rated as being at unclear risk of bias

  • Outcome includes studies that have been rated as being at high risk of bias in two or more categories

Inconsistency

  • Large heterogeneity based on the similarity of point estimates, statistical heterogeneity and I2 ≥50%

Imprecision

  • Large CIs when data are presented as standardised mean difference

  • Substantial heterogeneity (I2≥50%)

  • If a recommendation or clinical course of action would differ if the upper versus the lower boundary of the CI represented the truth

  • Sample size <400 within the meta-analysis for each variable

Indirectness

  • Use of surrogate outcomes

Publication bias

  • Asymmetric funnel plot