Table 1 Levels of evidence and Grades of recommendation

Levels of evidence

 I Evidence from at least one large randomised, controlled trial of good methodological quality (low potential for a bias) or meta-analyses of well-conducted randomised trials without heterogeneity

 II Small randomised trials or large randomised trials with a suspicion of bias (lower methodological quality) or meta-analyses of such trials or of trials with demonstrated heterogeneity

 III Prospective cohort studies

 IV Retrospective cohort studies or case–control studies

 V Studies without control group, case reports, and experts’ opinions

Grades of recommendation

 A Strong evidence for efficacy with a substantial clinical benefit, strongly recommended

 B Strong or moderate evidence for efficacy but with a limited clinical benefit, generally recommended

 C Insufficient evidence for efficacy or benefit does not outweigh the risk or the disadvantages (adverse events, costs,…), optional

 D Moderate evidence against efficacy or for adverse outcome, generally not recommended

 E Strong evidence against efficacy or for adverse outcome, never recommended