Precision of Optic Nerve Head and Retinal Nerve Fiber Layer Parameter Measurements by Spectral-domain Optical Coherence Tomography

PURPOSE: The aim of this study was to assess the repeatability and reproducibility (R&R) of Bruch membrane opening based on minimum rim width (BMO-MRW), minimum rim area (BMO-MRA) and peripapillary retinal nerve fiber layer thickness (RNFLT) with the Spectralis optical coherence tomography (Heidelberg Engineering) for normal and glaucoma subjects. Precise measurement of these parameters can support detection of structural glaucomatous damage and progression. METHODS: This cross-sectional study included 16 healthy controls and 16 patients with glaucoma. One eye was randomly selected and included in this study. Subjects underwent 1 baseline and 3 follow-up measurements, using 3 different Spectralis optical coherence tomography devices in randomized order, each operated by a single operator. Outcome measures were global and sectorial averages of BMO-MRW and BMO-MRA, and of peripapillary RNFLT obtained from 12/14/16-degree circle scans. Coefficients of variation (COV) were calculated and a mixed-effects analysis of variance was performed to compare R&R between devices. RESULTS: COVs of global and sectorial BMO-MRW measurement under repeatability conditions ranged from 0.51% to 1.7% (normal, 0.62% to 1.3%; glaucoma, 0.64% to 2.3%). Respective COVs under reproducibility conditions ranged from 0.89% to 1.9% (normal, 0.77% to 2.8%; glaucoma, 1.1% to 2.6%). COVs of global and sectorial RNFLT measurements under repeatability conditions ranged from 0.5% to 2.8%. Respective COVs under reproducibility conditions ranged from 1.6% to 3.5%. CONCLUSIONS: For R&R, the COVs of measured parameters were by trend higher for glaucoma eyes compared with normal controls. The BMO-MRW measurement system has an excellent precision taking into account that major and minor corrections of segmentation have to be done by the examiner before evaluation.