Reproducibility of peripapillary retinal nerve fiber layer thickness with spectral domain cirrus high-definition optical coherence tomography in normal eyes.

PURPOSE: To assess the reproducibility of the new spectral domain Cirrus high-definition optical coherence tomography (HD-OCT; Carl Zeiss Meditec, Dublin, CA, USA) for analysis of peripapillary retinal nerve fiber layer (RNFL) thickness in healthy eyes.
METHODS: Thirty healthy Korean volunteers were enrolled. Three optic disc cube 200 x 200 Cirrus HD-OCT scans were taken on the same day in discontinuous sessions by the same operator without using the repeat scan function. The reproducibility of the calculated RNFL thickness and probability code were determined by the intraclass correlation coefficient (ICC), coefficient of variation (CV), test-retest variability, and Fleiss' generalized kappa (kappa).
RESULTS: Thirty-six eyes were analyzed. For average RNFL thickness, the ICC was 0.970, CV was 2.38%, and test-retest variability was 4.5 microm. For all quadrants except the nasal, ICCs were 0.972 or higher and CVs were 4.26% or less. Overall test-retest variability ranged from 5.8 to 8.1 microm. The kappa value of probability codes for average RNFL thickness was 0.690. The kappa values of quadrants and clock-hour sectors were lower in the nasal areas than in other areas.
CONCLUSIONS: The reproducibility of Cirrus HD-OCT to analyze peripapillary RNFL thickness in healthy eyes was excellent compared with the previous reports for time domain Stratus OCT. For the calculated RNFL thickness and probability code, variability was relatively higher in the nasal area, and more careful analyses are needed.