Purchase this article with an account.
S. Dunker, A. U. Bayer, J. Sebag, German Eye Disease Study Group; German Eye Disease Study: Fundus Imaging Network Findings in Diabetic Retinopathy, Age-Related Macular Degeneration, and Systemic Hypertension. Invest. Ophthalmol. Vis. Sci. 2007;48(13):2400.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Within the German Eye Network, the Fundus Imaging Network (FIN) provides diagnostic services to 2 million people via 10 eye diagnostic centers, all with the same equipment, standardized testing paradigms, and centralized data processing. Since all patients are by referral, FIN is not a health screening system. This study evaluated the initial findings of this project to determine the frequency of major diseases and the ability of this approach to detect undiagnosed pathology.
Dilated fundus photography using the Zeiss FF450+ digital system generated 30° stereo central (optic disc and macula) as well as peripheral 50° digital pictures in each eye. Images were interpreted without knowledge of the history by 3 ophthalmologist graders following a study protocol. Each lesion was digitally recorded on each image and the totality of lesions for each eye was used to derive an ETDRS grading for diabetic retinopathy, an AREDS level for AMD, and a precise indication of vascular abnormalities such as focal narrowing, and arteriovenous nicking.
The reproducibility was 93% for different ophtalmologists grading the same patient. Of the first 243 patients evaluated, 31 (15.2%) had age-related macular degeneration (AMD), but only 3 (1.2%) were exudative. Hypertensive retinopathy was detected in 60/243 (24.7%), but only 41 (16.4%) were known to have systemic hypertension. Of another 235 eyes in patients with diabetes, 82 (34.9%) had ETDRS level 10, 98 (41.7%) had level 12, 31 (13.3%) level 15, and 6 (2.6%) had level 35, while 18 (7.6%) were indeterminate.
The reproducibility of FIN is high and the resulting data is useful for both clinical and research applications. The FIN approach has utility in detecting and characterizing fundus pathology that improves clinical care as in the case of hypertensive retinopathy, where one-third of patients with retino-vascular abnormalities were unaware that they had systemic hypertension. Due to the selection bias FIN is not an epidemiologic population-based project, and thus the distribution of diabetic retinopathy severities by ETDRS classification did not have the same prevalence and incidence of other studies. But the potential to rapidly accumulate large numbers of subjects evaluated in a standardized way and to follow them over time with the same methodologies makes FIN a powerful tool for clinical research, drug testing, and new diagnostic instrument development, as well as an effective way to improve clinical care.
This PDF is available to Subscribers Only