Agreement of New Automated Matched Alternation Flicker using Undilated Fundus Photography for the Detection of Glaucomatous Structural Change

Abstract

PURPOSE: To determine the agreement among glaucoma experts and general ophthalmologists regarding detection of glaucomatous structural changes using a new automated matched alternation flicker (AMAF) method with fundus photographs (FPs) of undilated eyes. METHODS: Sixty-six pairs of FPs of normal tension glaucoma patients were collected. FPs were taken at intervals of more than 12 months. Alternating flicker images were created using a new AMAF application. In a blinded manner, two glaucoma experts and two general ophthalmologists compared the presence of glaucomatous structural changes using either the AMAF method or the side-by-side comparison method. The interobserver and intraobserver agreements were compared using the Bland and Altman plot analysis. RESULTS: The glaucoma experts detected more glaucoma progression using the AMAF method (average, 50.7%) compared with the side-by-side method (average, 32.5%). General ophthalmologists detected more glaucomatous progression with the AMAF method (average, 40.9%) than with the side-by-side method (average, 25.0%). The AMAF method showed fair to substantial interreader agreement (k = 0.511-0.724) and fair to perfect intrareader agreement (k = 0.631-0.943). Interreader and intrareader agreements using the AMAF method were better for the glaucoma experts compared with the general ophthalmologists. CONCLUSIONS: The AMAF method showed more changes in FPs than the classical side-by-side comparison method. Regarding inter- and intrareader agreements, agreement for the glaucoma experts was best using the AMAF method, but for the general ophthalmologists agreement was best using the side-by-side comparison method.