“…Different methods were used to assess DR: ophthalmoscopy after mydriasis in four studies [21,22,24,30,36]; fundus photography after mydriasis in one study [27]; ophthalmoscopy without mydriasis in four studies [15,23,26,28,29,33,34,40]; and in nine studies no relevant details were provided [8, 10, 25, 31-33, 35, 38, 39]. Of the nine prospective studies assessed, five were screening studies conducted on consecutive patients with type 2 diabetes mellitus and proteinuria [21][22][23][24][25], while the remaining four were conducted on selected type 2 diabetes mellitus populations using criteria for the biopsy for type 1 diabetes mellitus (microhaematuria, absence of DR, atypical change in renal function, or immunological abnormalities) [26][27][28][29]. However, in most DR predicting DN Forest plots of the pooled sensitivity and specificity are shown in Fig.…”