Comparison of stereo disc photographs and alternation flicker using a novel matching technology for detecting glaucoma progression.

Abstract

Background and objective: To compare agreement of automated alternation flicker and serial stereophotograph inspection for detection of progressive glaucoma.

Patients and methods: Serial photographs of patients with glaucoma with at least 36 months of follow-up and perimetry every 4 months were assessed by four graders using predefined criteria with both flicker and stereophotography. The main outcome measure was progressive neuroretinal rim deterioration as identified by each technique.

Results: Forty eyes (20 patients) were included and 12 eyes progressed with perimetry. Using stereophotography, the overall agreement (kappa ± standard error) was 0.19 ± 0.06 for rim change, 0.78 ± 0.06 for disc hemorrhage, and -0.04 ± 0.06 for vessel movement. Using flicker, the overall agreement was similar for rim change (0.28 ± 0.06; P = .29), worse for disc hemorrhage (0.43 ± 0.06; P < .001), and better for vessel movement (0.22 ± 0.06; P = .002). The agreement between perimetric and disc progression was similar using stereophotography (0.10 ± 0.05) and flicker (0.19 ± 0.05; P = .20).

Conclusion: Agreement between flicker and stereophotography was similar.