As mentioned quite some time ago, there are a number of ways of approaching the ideas of indicating the strength of evidence behind recommendations. Archi has stuck with a rather old, but easy-to-follow version from the Centre for Evidence Based Medicine in Oxford. During the decade of Archi’s existence there’s been the steady development of a tool handily called GRADE. This produces high to very-low strengths of evidence, which translate into strong or weak recommendations. Explaining this seemed to be complicated, but Jeremy Howick has cracked a straightforward approach.
GRADE recommendations for interventions can be achieved by following a five-step process:
- Rank on the basis of randomised (high) or observational (low) study design
- Upgrade or downgrade on specific features of study results or conduct
Up for Consistent effects, Dose response or Confounders plausibly reducing effects seen
Down for High risk of bias in individual studies, publication or small study bias, Inconsistent, Indirect or Imprecise results
- Use this to assign an evidence level (High, Moderate, Low or Very Low)
- Consider other factors affecting a recommendation – like balances of harms and benefits, patient preferences and cost effectiveness
- Develop the strength of the recommendation (strong or weak)
Such decisions can be supported with standardised templates for the evidence, and as mentioned previously putting your recommendations in this frame can be very helpful to everyone.