Purpose: We evaluated the Glaucoma-Scope's ability to acquire images of the optic disc, the reproducibility of its analysis, and the comparison between it and expert evaluation.
Methods: One hundred and sixty-eight patients with satisfactory standard fundus photography were referred for Glaucoma-Scope analysis. Characteristics of eyes in which Glaucoma-Scope analysis could not be obtained were evaluated. Eight glaucoma patients (15 eyes) were quantitatively evaluated: inter- and intraobserver reproducibility of the Glaucoma-Scope analyses were calculated utilizing weighted Kappa index of agreement and coefficient of variation. The Glaucoma-Scope cup-to-disc ratio was calculated from the gray scale analysis and compared to the ratio as evaluated by a glaucoma expert.
Results: Glaucoma-Scope analysis was not achieved in 14% of a total of 336 eyes of 168 patients. The conditions associated with the inability to obtain a satisfactory image were hyperpigmented fundi, pseudophakia, aphakia, corneal opacities, cataract, and contact lenses. The quantitative analysis of the 15 patients showed small inter- and intraobserver means of the coefficient of variation (0.136 +/- 0.023 and 0.129 +/- 0.016) and SD (0.063 +/- 0.0056 and 0.064 +/- 0.0077) and moderate-substantial intra- and interobserver agreement (weighted K = 0.427 +/- 0.1 and 0.61 +/- 0.14, respectively). Significant disagreement (weighted K = -0.11, p = 0.0023) was found between Glaucoma-Scope (mean 0.52 +/- 0.064) and expert's evaluations (0.673) of the cup-to-disc ratio.
Conclusions: The inter- and intraobserver reliability and low variation of the Glaucoma-Scope readings indicates moderate to substantial reproducibility for detecting changes in optic disc's topography under selected conditions. However, eyes with moderate cataract, aphakia, pseudophakia, mild corneal opacity, darkly pigmented fundi or myopia were inconsistently analyzable. In a small group of randomly selected eyes there was very low agreement between the glaucoma expert's evaluation of the optic disc and that of the Glaucoma-Scope. These findings raise questions about the usefulness of the instrument as a diagnostic tool for glaucoma but show promise for detecting change when good images are possible.