Fig 11 Trial sequential analysis for retinopathy. Heterogeneity corrected required information size of 43 960 participants calculated on basis of proportion of retinopathy of 14.3% in conventional glucose control group, relative risk reduction of 10%, α=5%, β=20%, and I²=59%. Actually accrued No of participants was 10 793, 25% of required information size. Dashed red cumulative Z curve does not cross solid blue trial sequential monitoring boundaries for benefit or harm. Horizontal dotted green lines illustrate the traditional level of statistical significance (P=0.05)