To analyze the repeatability of measuring the thickness of the ganglion cell–inner plexiform layer using spectral-domain optical coherence tomography by auto-segmentation in various retinal diseases.

Test-retest reliability analysis.

A total of 180 patients who visited our clinic between April and December 2013 were included. An experienced examiner obtained 2 consecutive measurements from a macular cube 512 × 128 scan. The patients were divided into 3 groups according to central macular thickness. Repeatability was determined by comparing the average, minimum, and 6 sectoral ganglion cell–inner plexiform layer thickness values among groups.

Data from 165 of the 180 patients were included in the analysis. Measurement errors occurred in 3 patients (6%) in the macular atrophy group and 12 (20%) in the edema group. The measurement repeatability for ganglion cell–inner plexiform layer thickness was high in the normal eye group (average intraclass correlation coefficient [ICC], 0.994; coefficient of variation [COV], 0.7%; and test-retest variability, 1.1 μm), relatively low in the macular edema group (average ICC, 0.845; COV, 18.7%; and test-retest variability, 27.8 μm), and lowest in the macular atrophy group (ICC, 0.610; COV, 30.4%; and test-retest variability, 29.2 μm) compared to the normal eye group.

The repeatability of ganglion cell–inner plexiform layer thickness measurements was lower in the macular edema and atrophy groups. The impact of changes in macular shape caused by various retinal diseases should be taken into consideration when measuring ganglion cell–inner plexiform layer thickness in other diseases such as glaucoma and neuro-ophthalmology.