| Strength of Recommendation | |||
| Strong Recommendation = We Recommend | Weak Recommendation = We Suggest | ||
| Assessment Criteria for Certainty of Effect | |||
| Study Design |
Certainty of Effect Begins at This Level |
Lower if |
Higher if |
| Randomized trial |
High or moderate |
Risk of bias Inconsistency Indirectness Imprecision Publication bias |
Large effect Dose response All plausible confounding would reduce demonstrated effect or would suggest a spurious effect when results show no effect |
| Observational trial |
Low or very low |
||
Table 1: GRADE indicates Grading of Recommendations, Assessment, Development, and Evaluation.