7.7 Evidence-based recommendations for practice

The following is an extract from the 2009 guidelines of the Bundesaerzekammer (German Medical Association) www.arzt.de

Grading of recommendations

  • Level 1: based on available data, the benefits to the patient of complying with the recommendation are judged by experts to outweigh the potential risk
  • Level 2: if there are no definite data on the risk-benefit ratio

Grading of level of evidence

  • Level A: data from large, prospective, randomised studies
  • Level B: data from several prospective studies with conflicting results or with methodological flaws
  • Level C: data from case reports and non-randomised studies.
  • Level C+: data from case reports and non-randomised studies are unambiguous and confirmed by several investigations

Consequences of the recommendations

Both the level of evidence based on underlying data and the level of recommendation, reflecting the risk-benefit ratio impact on recommendation for medical practice (table 7.3).

Table 7.3 Classification of recommendations for clinical transfusion guidelines

Reproduced from: cross sectional guidelines for therapy with blood components and plasma derivatives, 4th revised edition 2009. Bundesaertztekammer (German Medical Association)

Level of recommendation Risk-benefit ratio Level of evidence Assessment of the methodological validity of the underlying data Overall assessment, classification Implications Key words
1 Unambiguous A Randomised, controlled studies without essential methodological flaws with unambiguous results 1 A Strong recommendation.
Valid for most patients
shall
1 Unambiguous C+ No randomised, controlled studies, but unambiguous data available 1 C+
1 Unambiguous B Randomised, controlled study with methodological flaws. Despite unambiguous results of the study, it cannot be safely ruled out that methodical flaws have influenced the results 1 B Strong recommendation.
Probably valid for most patients
1 Unambiguous C Observational studies without control group, but with convincing results 1 C Medium-strong recommendation,
seems to be plausible, may be changed once improved data becomes available
Should
2 Ambiguous A Randomised, controlled study without methodological reservations, but with conflicting results 2 A Medium-strong recommendation,
depending on the individual case, a different course of action may be indicated. The interpretation of results by the Working Group Guidelines are taken into account in the recommendation
2 Ambiguous C+ No randomised, controlled studies, but data can be extrapolated from other studies 2 C+ Weak recommendation,
depending on the individual case, a different course of action may be indicated. The interpretation of results by the Working Group Guidelines are taken into account in the recommendation
Can
2 Ambiguous B Randomised, controlled study with severe flaws 2 B Weak recommendation,
depending on the individual case, a different course of action may be indicated
Can
2 Ambiguous C Observational studies, case reports 2 C Very Weak recommendation,
depending on the individual case, a different course of action may be indicated
Could
Cross sectional guidelines for therapy with blood components and plasma derivatives, 4th revised edition 2009. Reproduced with permission from Bundesaertsekammer (German Medical Association).