Fig 1.
Flowchart of the 2-step study selection.
Fig 2.
Pooled sensitivity, specificity and diagnostic odds ratio of any DR.
Fig 3.
Summary receiver operating characteristic (SROC) curve.
Fig 4.
Positive and negative predictive values of DR for patients with diabetes type 1: (TP) true positive, (TN) true negative, (FP) false positive, (FN) false negative, (ppv) positive predictive value, (npv) negative predictive value.
ppv = TP/(TP+FP) = 73% of all positively tested diabetes patients, 73% have a true DR and 27% do not. npv = TN/(TN+FN) = 95% of all negatively tested diabetes patients, 95% are truly negative and 5% are not.
Fig 5.
Positive and negative predictive values of DR for patients with diabetes type 2: (TP) true positive, (TN) true negative, (FP) false positive, (FN) false negative, (ppv) positive predictive value, (npv) negative predictive value.
ppv = TP/(TP+FP) = 49% of all positively tested diabetes patients, 49% have a true DR and 51% do not. npv = TN/(TN+FN) = 98% of all negatively tested diabetes patients, 98% are truly negative for DR, while 2% are not.
Fig 6.
Funnel plot to assess publication bias.