Purchase this article with an account.
A. Mishra, P. H. Artes, G. P. Sharpe, M. T. Nicolela; Sensitivity of Spectralis OCT Circumpapillary Retinal Nerve Fibre Layer Thickness Measurements and Relationships With Visual Field Damage in Three Distinct Types of Morphological Disc Damage in Glaucoma. Invest. Ophthalmol. Vis. Sci. 2010;51(13):4915.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To compare diagnostic sensitivity of spectral-domain OCT circumpapillary RNFL thickness measurements and structure-function relationships in patients with three distinct types of morphological optic disc damage in glaucoma.
A group of 84 patients enrolled in a prospective study on progression with different types of glaucomatous optic disc damage (27 with focal damage [MD, -7.6 ± 5.0 dB], 30 with diffuse damage [MD, -4.4 ± 4.5 dB] and 27 with sclerotic damage [MD, -3.7 ± 3.6 dB]) underwent circumpapillary OCT scans of the retinal nerve fibre layer (RNFL) with a spectral-domain OCT (Spectralis, Heidelberg Engineering, HRAViewer 184.108.40.206). The last 3 visual fields (Humphrey Field Analyzer, program 24-2, SITA-Standard) obtained within 12 months were averaged to obtain an accurate estimate of visual function. Healthy controls (n=58) were examined to estimate specificity.
At specificities of 95% (81% when borderline classifications were treated as test-positives), the diagnostic sensitivity of the circumpapillary RNFL scan was 93% (100%) in patients with focal damage, 70% (80%) with diffuse damage, and 76% (79%) with sclerotic damage. While the correlation between global measures of RNFL thickness and visual field damage (MD) was only modest (Spearman r = 0.40, p<0.001), stronger relationships were apparent with more localized analyses, particularly between the sector MD of the superior field and the inferior RNFL, Spearman r = 0.58). However, systematic differences between the structure-function relationships of the three morphological groups were not apparent.
The diagnostic accuracy of the circumpapillary RNFL protocol is high. However, the large residual scatter makes it difficult to determine differences between the structure-function relationships of subgroups with distinct morphological damage. Our data confirm prior observations by Hood et al that this scatter is chiefly due to systematic variation between individuals rather than measurement error.
This PDF is available to Subscribers Only