A comparison between one- and two-field 60°fundus photography when screening for diabetic retinopathy

Abstract

Purpose: To compare the severity level of diabetic retinopathy obtained when assessed from two versus only one 60°photographic field using colour transparencies and red-free, black-and-white photographs. To compare the areal coverage of these two photographic strategies to that of seven-field 30°photography. Methods: Two ophthalmologists graded photographs of 74 eyes of 74 type I and II diabetes patients. Inter-method agreement was expressed in percentages and using kappa statistics and scatter-diagrams. The comparison of the approximate photographic areal coverage was done from diagrams using planimetry. Results: The severity level of retinopathy when judged from two photographic fields was more severe in 13.5% (Grader 1) and in 16.2% (Grader 2) from colour transparencies and in 13.5% (Grader 1) and in 14.9% (Grader 2) from red-free black-and-white prints, as compared to assessments from only one field. Kappa values (0.84-0.86) for inter-method agreement for five pooled retinopathy levels revealed good agreement. Neither grader missed retinopathy requiring clinical assessment or treatment (levels ≥47) when minimal retinopathy (levels 14-20) was detected using only one 60°colour slide or red-free photograph. A second optic disc-centred field provided valuable additional information when more severe retinopathy lesions (levels ≥30) were detected in the macula-centred field. One macula-centred 60°photograph covered 60% and two 60° photographs 80% of the area covered by seven-field 30°photography. Two-field 60°photography covers areas left outside seven-field 30°photography. Conclusion: We propose the use of one macula-centred 60°photograph when screening for the first lesions of diabetic retinopathy. After they have been found two-field 60°photography is recommended.