Xuxi Zhang

Supplementary Table S7 GRADE Evidence Profile: DR Progression Quality assessment No of patients Effect Quality Importance No of studies Design Risk of bias Inconsistency Indirectness Imprecision Other considerations Interventions targeting modifiable risk factors of DR Control Relative (95% CI) Absolute DR Progression (follow ‐ up 3.8 to 8 years) 10 randomised trials no serious risk of bias very serious 1 no serious indirectness no serious imprecision none 747/9584 (7.8%) 862/9483 (9.1%) OR 0.74 (0.59 to 0.92) 22 fewer per 1000 (from 7 fewer to 35 fewer)  LOW CRITICAL DR Progression: Blood ‐ pressure ‐ control intervention (follow ‐ up 4 to 8 years) 2 randomised trials no serious risk of bias no serious inconsistency no serious indirectness serious 2 none 67/645 (10.4%) 64/616 (10.4%) OR 1.05 (0.77 to 1.45) 5 more per 1000 (from 22 fewer to 40 more)  MODERATE IMPORTANT DR Progression: Glycemic ‐ control intervention (follow ‐ up 4 to 8 years) 4 randomised trials no serious risk of bias very serious 1 no serious indirectness no serious imprecision none 524/7592 (6.9%) 597/7578 (7.9%) OR 0.71 (0.52 to 0.97) 22 fewer per 1000 (from 2 fewer to 36 fewer)  LOW IMPORTANT DR Progression: Lipid ‐ control intervention (follow ‐ up 4 to 8 years) 2 randomised trials no serious risk of bias very serious 1 no serious indirectness serious 2 none 99/1205 (8.2%) 117/1150 (10.2%) OR 0.83 (0.44 to 1.59) 16 fewer per 1000 (from 54 fewer to 51 more)  VERY LOW IMPORTANT DR Progression: Multifactorial intervention (follow ‐ up 3.8 years) 2 randomised trials no serious risk of no serious inconsistency no serious indirectness serious 2 none 57/140 (40.7%) 84/139 (60.4%) OR 0.39 (0.23 to 0.65) 231 fewer per 1000 (from 106  MODERATE IMPORTANT 7 199 Prevention and control of diabetic retinopathy