Purchase this article with an account.
Scott D. Smith, Nathan M. Radcliffe, Zeba A. Syed, Gustavo V. De Moraes, Sung Chul Park, Jeffrey M. Liebmann, Robert Ritch; Detection Of Glaucoma Progression By Traditional And Alternation Flicker Review Of Stereo Disc Photographs. Invest. Ophthalmol. Vis. Sci. 2011;52(14):4168.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To compare the detection and inter-observer agreement in grading of glaucomatous optic disc progression using automated alternation flicker and side-by-side (SBS) viewing of stereo optic disc photographs.
Serial stereophotographs from 125 consecutive glaucoma patients with >8 reliable visual fields (VF) were obtained. Eyes were categorized as having rapid, moderate, or slow VF change by the global rate of threshold sensitivity change. Two graders masked to progression status reviewed the stereophotographs in random order both with SBS viewing and aligned with automated alternation flicker (EyeIC, Narbeth, PA). Each grader identified disc hemorrhage (DH), parapapillary atrophy (PPA) change, vessel movement (VM), rim change, and retinal nerve fiber layer (RNFL) change. Inter-observer agreement for each parameter and overall progression based on change in any parameter was assessed using the kappa statistic. The proportion of eyes classified as having disc progression by each viewing method was also compared for each grader, stratified by the rate of visual field progression.
Inter-observer agreement was greater using flicker than SBS review for DH (kappa: 0.75 vs 0.67), PPA change (0.57 vs 0.41), VM (0.62 vs 0.33), rim change (0.59 vs 0.30) and RNFL change (0.56 vs 0.39). Agreement for overall assessment of disc progression was also greater using flicker (0.70 vs 0.39). Detection rates for grader 1 were significantly higher (all p≤0.0001) using flicker than SBS viewing for DH (12.0% vs 5.1%), PPA (22.8% vs 6.3%), VM (16.4% vs 2.5%), and rim change (17.1% vs 3.2%). Detection rates for grader 2 were significantly higher (all p<0.02) using flicker for PPA change (15.8% vs 5.1%) and VM (15.2% vs 8.2%). Rates for identifying overall progression were higher using flicker for both grader 1 (35.4% vs 10.8%) and grader 2 (31.0% vs 20.9%; both p<0.00005). For each grader, more individual parameters and the classification of overall progression were more strongly associated with the rate of VF change using flicker than SBS viewing.
Inter-observer agreement and sensitivity of detection of optic disc progression are increased by review of stereophotographs using alternation flicker.
This PDF is available to Subscribers Only