Strategies of digital fundus photography for screening diabetic retinopathy in a diabetic population in urban China.

Jiyuan Ding,Yanhong Zou,Ningpu Liu,Li Jiang,Xuetao Ren,Wei Jia,Torkel Snellingen,Virasakdi Chongsuvivatwong,Xipu Liu

DOI: https://doi.org/10.3109/09286586.2012.716895

2012-01-01

Ophthalmic Epidemiology

Abstract:Purpose: To evaluate the effect of mydriasis and different field strategies on technical failure, probability to refer diabetic retinopathy (DR, sensitivity) and probability not to refer patients without DR (specificity) of digital photography in screening with a fundus camera. Methods: A total of 531 patients with diabetes underwent fundus photography with cross-combinations of mydriasis/nonmydriasis and single-field/two-field strategies, followed by slit lamp biomicroscopic examination by a trained ophthalmologist. Fundus photographs were graded independently by another experienced ophthalmologist. Calculations were first based on cases with non-gradable images treated as being referred and then with them excluded. Results: Percentages of DR and referable DR in this patient cohort were 22.4% and 7.7%, respectively, based on slit lamp biomicroscopic examination. Mydriasis significantly reduced the technical failure rate from 27.1% to 8.3% under a single-field strategy and from 28.2% to 8.9% under a two-field strategy. As compared to the single-field strategy, the two-field strategy increased sensitivity from 75.6% to 87.8% without mydriasis and from 73.2% to 90.2% with mydriasis. Mydriasis increased specificity from 68.8% to 84.3% in the single-field strategy and from 64.7% to 81.6% in the two-field strategy. Had the subjects with non-gradable images been excluded, the two-field strategy without mydriasis reported sensitivity of 85.7% and specificity of 91.6%. Conclusions: Both mydriasis and the two-field strategy are useful in photographic screening tests. Technical failure should be taken into consideration when screening strategies for DR are determined.