The GRADE methodology for guideline development

A summary of the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) methodology will be presented. GRADE is a system of rating quality of evidence and grading strength of recommendations in systematic reviews, health technology assessments, and clinical practice guidelines addressing alternative management options. GRADE requires a clear specification of the relevant setting, population, intervention, and comparator for each management question. It also requires specification of the relative importance of the outcomes considered. GRADE then evaluates the quality of evidence on five specific domains: study limitations; inconsistency of results; indirectness of the evidence; imprecision and publication bias. Evidence generated from randomized control trials starts as high quality but can be downgraded if any of the above limitations are present. The quality of the evidence could be rated as high, moderate, low and very low. Guideline developers should make an overall rating of confidence (quality of evidence) in estimates of effect for each important or critical outcome. Recommendations are then characterized as strong or weak according to the balance between desirable and undesirable consequences of the alternative management options, the quality of the supporting evidence, estimates of values and preferences, and resource use. Ultimately, guideline panels must use judgment in integrating these factors to make a strong or weak recommendation The use of this system has several advantages; it uses a transparent and explicit approach and is evidence based. GRADE is now the preferred quality of evidence rating system, and has been adopted by over 50 organizations worldwide.