Reproducibility of optic nerve head measurements obtained by optical coherence tomography

Abstract

Purpose: To evaluate reproducibility of optic nerve head (ONH) morphometry measurements obtained by optical coherence tomography (the fast optical disk protocol of the -Zeiss model 3000 OCT system) in normal and glaucomatous eyes.

Methods: Prospective instrument-evaluation study. ONH measurements for 20 eyes were obtained in eight scan sessions taken during two visits to an ophthalmology clinic (10 normal patients, 10 glaucoma patients, one eye per subject). At every one of the eight sessions for each eye, estimates of eight ONH morphometry variables (see Main outcome measures) were obtained. The first two sessions were performed by two operators, followed by a 30-minute break. The same operators then completed a third and fourth session. This sequence was duplicated on a second visit. Intrasession, intersession, intervisit, and interoperator reproducibility of the eight variables were calculated by the use of a components variance model. Intraclass correlation coefficients (ICC) were used to assess reliability.

Results: Vertical integrated rim area, horizontal integrated rim area, disk area, cup area, rim area, cup/disk area ratio, cup/disk horizontal ratio, cup/disk vertical ratio. With the exception of the horizontal integrated rim area and rim area in normal subjects, the factor subject was the most important source of variance for all variables. Reliability values as measured by ICC for normal eyes were above 81%, with the exception of measurements of the horizontal integrated rim area (23.1%), rim area (33.3%), and disk area (64.7%). For glaucomatous eyes all values were above 85%, with the exception of the disk area (68.1%).

Conclusions: ONH measurements obtained using the fast optical disk protocol of the Zeiss 3000 OCT system show good reproducibility, for both normal and glaucomatous eyes.