Abstract
Purpose::
To determine the reproducibility of optical coherence tomography (OCT) measurements of macular, nerve fiber layer (NFL) and optic disc parameters, and to investigate the influences of axial length and refraction on these measurements in children with healthy eyes.
Methods::
The Sydney Childhood Eye Study is a population-based survey of 2353 Year 7 students (75.3% response) from a random cluster-sample of 21 secondary schools across Sydney. A consecutive sub-sample of 120 children had OCT (StratusOCT, Carl Zeiss, CA) performed by a single operator, which was repeated with a brief rest between the two sessions. Scans of the macula, NFL and optic disc were performed.
Results::
The mean age, refractive error and axial length of the children was 13.3 yrs, -0.06D and 23.36 mm, respectively. The inter-session intra-class correlation coefficients (ICC) for measurements of macular, NFL and disc parameters were small. Among all macular parameters, foveal minimum thickness was the least reproducible measure and also had the highest intra-session variability (ICC 14.4%) and coefficient of variation (CoV, 4.1% & 3.4%). Among NFL parameters, nasal quadrant measures were the least reproducible with relatively high intra-session ICC & CoV (5.9% & 7.2%). The reproducibility of optic disc measures was similar to NFL measures except for horizontal & vertical disc diameters, which showed low inter-subject ICC (48.4% & 38.6%, respectively) and high CoV (7.6% & 10.1%, respectively). There was strong agreement (correlation coefficients >0.7, p<0.0001) between measurements performed either with or without axial length and refraction data entered for all 3 comparisons. Absolute differences in measurements of macular parameters were not significantly different between scans performed with or without ocular biometric data entered. For NFL thickness measurements, mean differences in nasal and temporal quadrants were significant when the refractive error was entered into the instrument. There was no significant difference whether either axial length or both axial length and refraction were entered.
Conclusions::
The StratusOCT demonstrated quite reproducible measurements of macular and NFL thickness. Optic disc parameters, however, showed lower repeatability. Magnification due to axial length and refractive error had minimal impact on the measurements of macular and NFL thickness.
Keywords: macula/fovea • nerve fiber layer • optic disc