Xuxi Zhang

fewer to 255 fewer) New Onset DR: Follow ‐ up<2 years (follow ‐ up 1 to 2 years) 3 randomised trials serious 6 no serious inconsistency no serious indirectness no serious imprecision none 13/80 (16.3%) 17/68 (25%) OR 0.59 (0.15 to 2.34) 86 fewer per 1000 (from 202 fewer to 188 more)  MODERATE IMPORTANT New Onset DR: Follow ‐ up 2 ‐ 5 years (follow ‐ up 4 to 5 years) 4 randomised trials no serious risk of bias 5 no serious inconsistency no serious indirectness no serious imprecision none 100/312 (32.1%) 131/319 (41.1%) OR 0.59 (0.34 to 1.02) 119 fewer per 1000 (from 219 fewer to 5 more)  HIGH IMPORTANT New Onset DR: Follow ‐ up >5 years (follow ‐ up 6 to 7.5 years) 3 randomised trials no serious risk of bias no serious inconsistency no serious indirectness no serious imprecision none 105/2595 (4%) 104/2456 (4.2%) OR 0.57 (0.42 to 0.78) 18 fewer per 1000 (from 9 fewer to 24 fewer)  HIGH IMPORTANT 1 Four studies had high risk of performance bias [Kumamoto, 1995; Tovi, 1998; Rachmani, R., 2002; VA CSDM, 1996] 2 11 RCTs 3 Only two studies 4 Three studies had high risk of performance bias [Kumamoto, 1995; Tovi, 1998; VA CSDM, 1996] 5 One study had high risk of performance bias [Rachmani, R., 2002] 6 Two studies had high risk of performance bias [Tovi, 1998; VA CSDM, 1996] VA CSDM, 1996 DR, diabetic retinopathy 196 Chapter 7