RESUMO
BACKGROUND: The KDIGO (Kidney Disease: Improving Global Outcomes) clinical practice guidelines establish international recommendations for the definition and treatment of kidney disease. Our objective was to characterize the strength of evidence supporting the KDIGO guidelines, the class of recommendations made, and the relationship between these. STUDY DESIGN: We reviewed and abstracted the level of evidence and strength of recommendations in the currently available KDIGO guidelines. SETTING & POPULATION: KDIGO clinical practice guidelines target care of patients with kidney disease to improve outcomes. SELECTION CRITERIA FOR STUDIES: All KDIGO guidelines published on the KDIGO website as of November 2013 were included. PREDICTOR: Recommendations pertaining to disease, diagnosis, or treatment. OUTCOMES: Levels of evidence and strength of recommendations. RESULTS: Of 853 recommendations in 9 guidelines, 5% were supported by level A quality evidence; 17%, level B; 31%, level C; 18%, level D; and 20%, ungraded evidence. The strength of recommendations was class 1 for 25%, class 2 for 54%, and ungraded for 20%. Only 3% of recommendations were class 1 in strength and supported by level A evidence. Of the recommendations, 2% concerned disease definition and classification; 29%, diagnosis; and 69%, treatment. LIMITATIONS: Our study included only the KDIGO guidelines. We did not assess historical changes in nephrology guidelines recommendations. CONCLUSIONS: KDIGO recommendations were based largely on weak evidence, reflecting expert opinion. Few recommendations were both strong and supported by high-level evidence.