Purpose: To determine agreement between slit-lamp indirect ophthalmoscopy and Stratus optical coherence tomography (OCT) when assessing cup-to-disc ratios (CDRs).
Methods: Twenty-five ocular hypertensive subjects and 56 patients with primary open-angle glaucoma were included. Estimation of vertical (VCDR) and horizontal (HCDR) cup-to-disc ratio with slit-lamp ophthalmoscopy was made by three glaucoma specialists along with OCT scanning of optic nerve head. Agreement between OCT and specialists was measured by intraclass correlation coefficients (ICC), Bland and Altman's scatterplots, and a regression coefficient of the average difference.
Results: The mean VCDR and HCDR was significantly higher (P<0.001) with OCT than that estimated by the specialists, with the difference ranging from 0.08 to 0.11, and from 0.13 to 0.18, respectively, depending on the specialist. Difference was higher (P<0.001) for cuppings below 0.3, and looses significance for larger VCDR cuppings (above 0.7). ICC for VCDR was 0.87 among specialists, and ranges from 0.82 to 0.75 when comparing OCT and specialists. ICC for HCDR was 0.83 among specialists and 0.74 between OCT and specialists. When data were plotted according to the Bland-Altman method, as the cupping increased, the agreement also increased.
Conclusions: There is very good agreement among the specialists when estimating CDRs by stereoscopic slit-lamp biomicroscopy. OCT shows higher values than the specialists; the greatest differences occurred when assessing small CDRs and the differences diminished as the cupping increased. These two methods of measurement are not interchangeable, and the difference must be considered, especially in discs with smaller CDRs.