Detection of retinal neovascularizations using 45° and 60° photographic fields with varying 45° fields simulated on a 60° photograph

Abstract

Purpose: To compare the effectiveness of one- and two-field 45° and 60° fundus photography for detection of retinal neovascularizations (NVEs); to compare the detection of NVEs in macular 45° fields centred on the fovea, between the fovea and the disc and with the nasal border along the nasal disc margin with that in standard 60° one- and two-fields. Methods: Two ophthalmologists assessed red-free, black and white screening and follow-up photographs of 58 out of a total 72 patients referred for treatment for NVEs over the 3 year period 1997-99. The assessment included analysis of the numbers, sizes and locations of NVEs identified in different photographic fields defined by a transparent grid placed over two standard 60° fields. Results: Testing of 45° macular fields disclosed 59/81 (73%), 50/81 (62%) and 64/81 (79%), respectively, of NVEs detected in a 60° fovea-centred photograph. Kappa values for detecting at least one NVE, when present, were 0.62 (95% CI 0.40-0.83), 0.38 (95% CI 0.19-0.57) and 0.83 (95% CI 0.62-0.99). The corresponding figures when compared with two-field 60° photography were 59/111 (53%), 50/111 (45%) and 64/111 (58%), respectively. One macula-centred 60° field disclosed 81/111 (73%) of NVEs detected by 60° two-field photography with an additional optic disc-centred field. Two-field 45° photography disclosed 86/111 (77%) of NVEs detected by two-field 60° photography. A total of 90% of NVEs detected in the optic disc-centred 60° field only were also detected in the corresponding 45° field. Conclusion: One-field 45° photography cannot be recommended as a screening method for diabetic retinopathy. Two-field 45° photography (macula-centred and disc-centred fields) disclosed 77% of NVEs detected by the corresponding two-field 60° photography.