Purchase this article with an account.
J. Sherman, E. Laskova, M. Battaglia; Sensitivity, Specificity and Reproducibility of GDx VCC, ECC and Screening Program . Invest. Ophthalmol. Vis. Sci. 2005;46(13):2524.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Purpose: To assess the sensitivity, specificity and reproducibility of the GDx VCC (previously reported) and compare it to both the new GDx ECC (extended corneal compensator or " bias" method) and the new screening program (both not previously reported). Methods: As part of a multi–center study to create a new database for the GDx ECC, 45 eyes from 27 subjects with glaucoma (abnormal exam + reproducible glaucomatous–type field defect on SITA standard 24–2 V.F.) and 40 eyes from 20 normal subjects (normal exam+normal 24–2 SITA standard V.F.) were evaluated. Each subject had 3 VCC and 3 ECC scans and at least 1 screening on the same day. Classification of abnormal GDx results are based upon an NFI above 30 or superior average or inferior average or TSNIT average flagged at the p< 1% level (color code red or yellow). The Screening mode(SCC) classified the results into three parameters: Within normal limits, Borderline and Outside Normal Limits. Only results that fell Within normal limits were classified as normal. Results that fell within Bordeline and Outside Normal Limits were classified as abnormal. Reproducibility for three GDx parameters (TSNIT, Superior average and Inferior Average) between the three scans was calculated. Results: Of the 45 eyes with glaucoma, 39 had an abnormal GDx VCC, ECC and SCC (sensitivity of all three = 87%). Perfect agreement was not found, however, in the 9 "normal" eyes of the glaucoma subjects. Of the 40 normal eyes, 40 had a normal VCC, ECC and SCC (specificity of all three = 100%).With regard to classification, there was no difference between VCC and ECC. However, ECC did improve the occasional spurious VCC findings such as the "tie–dye" pattern. For three trials in each eye in 27 glaucoma subjects, the variation average, the averaged standard deviation (µm) and 95% confidence interval range, respectively, were: VCC: 0.84µm, 1.24µm and (0.65,1.03)µm; ECC: 1.04µm, 0.84µm and (0.90,1.17)µm. For three trials in each eye in 20 normal subjects, the variation average, the averaged standard deviation and 95% confidence interval range, respectively, were: VCC: 0.86µm, 0.61µm and (0.75,0.97)µm; ECC: 1.17µm, 0.87µm and (1.01,1.33)µm. Conclusions: GDx VCC, ECC and SCC all have been shown to have equal and high sensitivity and specificity (87% and 100% ). Both VCC and ECC are highly reproducible but VCC measurements appear to have a slightly higher reproducibility than ECC measurements. However, the ECC demonstrates fewer artifacts such as the "tie–dye" pattern. SCC reproducibility studies are recommended.
This PDF is available to Subscribers Only