Purchase this article with an account.
JN Kirkpatrick, P Scanlon, R Malhotra, B Harney, G Thomas, C Foy, N Lewis Barned; Efficacy of Digital Diabetic Retinopathy Screening : A Population Based Survey . Invest. Ophthalmol. Vis. Sci. 2002;43(13):4387.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Purpose: To evaluate the effectiveness of digital fundus photography as a method for diagnosis of diabetic retinopathy and for detection of sight threatening retinopathy. Methods: 3611 patients were selected randomly from the diabetic population of Gloucestershire (c 12,000 patients) undergoing systematic retinopathy screening. All patients had digital fundus photography according to the Eurodiab protocol, with and without mydriasis, using a Topcon TRC NW55 camera with a Sony 3-chip video attachment. Of these, 1549 patients underwent simultaneous slit lamp biomicroscopy by an experienced ophthalmologist. All images were graded for presence of diabetic retinopathy using a modified protocol based on the ETDRS grading system by experienced ophthalmologists. Results: With mydriasis sensitivity and specificity for the detection of STDR were 87.2% (CI 82.3 - 92.1%) and 90.1% (CI 88.5 - 91.7%) respectively with a technical failure rate of 4.4%. Absence of mydriasis reduced sensitivity and specificity to 76.5% (CI 70.3-82.7%) and 87.4% (CI 85.6 - 89.2%) respectively with a technical failure rate of 20.7%. Results comparing the grades of retinopathy detected by digital photography with gold standard slit lamp examination showed close agreement for all grades of retinopathy greater than mild NPDR. There was a relative underdiagnosis of normal eyes with digital imaging (50.2% versus 70.3%) and an overdiagnosis of minimal NPDR (14.9% versus 3.6%). Conclusion: Digital fundus photography is a convenient and acceptable method of screening for diabetic retinopathy with an efficacy that compares well to other screening methods. The resolution of digital imaging techniques as used in this study is insufficient to accurately detect minimal diabetic retinopathy.
This PDF is available to Subscribers Only