Influence of examiner experience on the reproducibility of retinal nerve fiber thickness values using Cirrus and Stratus OCTs.

PURPOSE: The purpose of the study was to evaluate the influence of the examiner's experience on the reproducibility of retinal nerve fiber layer (RNFL) measurements obtained with Cirrus optical coherence tomography (OCT) and Stratus.

METHODS: Fifty-one normal and glaucomatous eyes of 51 participants were included. Two examiners (1 novice, 1 experienced) obtained 2 scans using both OCTs. For quantitative measurements, Bland and Altman limits of agreement were evaluated. For qualitative classifications, κ coefficients were calculated.

RESULTS: Signal strength was higher with Cirrus than with Stratus (P<0.05). Signal strength was higher in scans performed by the experienced examiner than in those carried out by the inexperienced examiner in Stratus but not in Cirrus. RNFL measurement differences between Cirrus and Stratus were influenced by the examiner for the inferior (P=0.02), superior (P<0.001), and temporal quadrants (P=0.009). The RNFL quantitative agreement of examiners was higher in Cirrus than in Stratus. The qualitative agreement (κ coefficients) of both examiners in the RNFL classification were almost perfect with Cirrus (in the average, superior, and inferior quadrants), and moderate with Stratus (only in average and inferior quadrant).

CONCLUSIONS: The signal strength is independent of the examiner's experience in Cirrus but not in Stratus. RNFL measurements obtained by both examiners were more reproducible with Cirrus than with Stratus. The differences in RNFL measurements between both OCTs were related to the examiner's experience in all 3 quadrants. Agreement between operators in the RNFL classification was higher with Cirrus than Stratus.