Limits of two-dimensional planimetry in the follow-up of glaucomatous optic discs

To test the intra-observer, intra-photographic variability of two-dimensional measurements of the optic nerve head we used computer-assisted planimetry. Depending on the variability, we calculated the confidence intervals of the optic disc parameters which could be indicative of glaucomatous damage on follow-up.

Slides of the optic disc were taken from 10 eyes of 10 patients (n = 6 open angle glaucoma, n = 4 ocular hypertension) using a Zeiss fundus camera. All eyes were evaluated 10 times within a random sequence on 10 different days. We obtained the absolute values of the disc radii and the cup radii in steps of 1, 10, and 45 deg in predefined quadrants and the mean radii.

The confidence interval of the cup radius on follow-up, depending on sector size, ranged between 62 and 38% for small cups (radius 0.2 mm) and between 12 and 7% for large cups (radius 0.8 mm). The confidence intervals of the cup/disc ratio distinguishable from the disc boundary, depending on sector size, ranged between 0.81 and 0.89 for small discs (radius 0.5 mm) and from 0.90 to 0.94 for large discs (radius 1.0 mm). The confidence intervals of the cup/disc ratio indicating an increase on the cup radius in follow-up, distinguishable from the boundary of the disc, ranged, depending on sector size, between 0.57 and 0.75 for small discs (radius 0.5 mm) and from 0.81 to 0.89 for large discs (radius 1.0 mm).

The smaller the disc, the more difficult is the detection of glaucomatous damage, and the larger the cup, the more difficult is the detection of progression of glaucomatous damage.