Agreement between clinicians and a confocal scanning laser ophthalmoscope in estimating cup/disk ratios.

Abstract

We assessed agreement between cup/disk ratio measurements obtained by glaucoma expert evaluation of stereoscopic photographs of the optic disk and those obtained with a confocal scanning laser ophthalmoscope. Three glaucoma experts estimated vertical and horizontal cup/disk ratios from stereoscopic photographs of 15 normal subjects and 15 patients with glaucoma. These estimates were compared to vertical, horizontal, and area cup/disk ratios measured with a confocal scanning laser ophthalmoscope. Intraobserver and interobserver agreements were also estimated. Agreement between clinicians and the confocal scanning laser ophthalmoscope varied by clinician. Agreement was moderate to substantial for vertical cup/disk ratio and fair to moderate for horizontal cup/disk ratio; kappas ranged from 0.57 to 0.72 and from 0.21 to 0.55, respectively. The mean confocal scanning laser ophthalmoscope area cup/disk ratio measurements were smaller than each clinician's mean vertical and horizontal cup/disk ratio estimates; differences ranged from 0.10 to 0.24 and from 0.06 to 0.16, respectively. Differences were smaller between clinician estimates and instrument measurements of horizontal and vertical cup/disk ratios of patients with glaucoma than normal subjects. These results demonstrate good agreement between confocal scanning laser ophthalmoscope measurements and clinician estimates of the vertical cup/disk ratios from stereoscopic photographs, particularly of patients with glaucoma. However, as differences between clinician and instrument estimates of cup/disk ratios were found, new quantitative criteria must be established for characterizing a disk as glaucomatous using confocal scanning laser ophthalmoscopy.