Reproducibility of the OCT-3: Evaluation of Glaucoma Variables

Abstract

Abstract: : The OCT-3 is a commercial imaging device with algorithms for evaluating the retina and optic nerve head. Purpose: to determine the intra and inter observer variablity of glaucoma measurements. Methods: 20 subjects (10 normal,10 glaucoma) were examined using fast NFL and fast ONH analysis 5X by an examiner followed by a second examiner repeating the same tests in random order. Subjects were dilated and underwent the same tests by the same examiners in random order. Repeat examinations were performed 1- 2 weeks later. Results: 30% of normal tests could not be obtained without dilation; undilated examinations were abandoned for glaucomatous eyes. Shrout Fleiss correlation coefficients, an indicator of reliability were calculated 1) Interobserver ONH parameters correlations range from: 0.87 – 0.96 for dilated and 0.54 – 0.59 for undilated normals and from 0.88 -0.97 for glaucomas 2) Intraobserver ONH 0.90-0.97 dilated, 0.40-0.68 undilated normals and 0.92-0.98 in glaucomas 3) Interobserver RNFL parameters 0.90-0.97 normal dilated, 0.40-0.91 undilated and 0.85-0.97. glaucoma 4) Intraobserver 0.75-0.97 dilated normal, 0.64-0.88 undilated and 0.82-0.97 in glaucoma. Conclusions: Overall reproducibility was excellent in dilated eyes and fair to good in undilated eyes. NFL thickness was less reproducibile nasally and temporally than in the polar regions.

Keywords: imaging/image analysis: clinical • nerve fiber layer • optic disc