Purchase this article with an account.
R. Bourkiza, V. Vora, G. Vafidis; Is an Additional Temporal Field Image Useful to Grade Disease Activity in Treated Proliferative Diabetic Retinopathy?. Invest. Ophthalmol. Vis. Sci. 2010;51(13):4683.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
In the English National Screening Programme for Diabetic Retinopathy (ENSPDR) two standardised 45° digital images of each eye are graded to determine retinopathy status and outcome pathway. Treated proliferative diabetic retinopathy (PDR) is graded on current disease activity. Since clinical examination temporal to the macula may reveal disease activity not apparent elsewhere, it has been our policy since January 2007 to take an additional temporal image. The study analysed the diagnostic value of this image.
We reviewed the hospital DR image database to identify patients with treated PDR and an additional temporal field. Using high resolution computer screen with magnification and red-green image manipulation, randomly ordered standard image sets were graded twice by an accredited masked ENSPDR grader: first without, then with the additional temporal image. Lesions were graded as background (R1), pre-proliferative (R2) or neovascular (R3) DR. The two grades were compared using the Wilcoxon matched-pairs signed rank test.
Of 4478 digital images taken over 3 years, 194 image sets (4.3%) had treated PDR with an additional temporal image (387 eyes). Of these 387 eyes, 100 (25.8%) were graded R3 (active proliferative disease). In 44 eyes (44%), new vessels were seen in the additional temporal image, and in 29 (29%) uniquely so. Of the remainder eyes with no evidence of neovascularisation in the extra-temporal field (n=343), further retinal photocoagulation was given to 28 eyes (8.1%) on the basis of pre-proliferative activity (R2) in the additional image. Overall, the additional image changed retinopathy grade in 15.2% (n=59) of eyes (z=-6.62, p<0.001 Two-Tail).
Although this is retrospective with possible selection bias, the results appear to support taking an additional temporal digital image to grade disease activity in treated PDR.
This PDF is available to Subscribers Only