Purchase this article with an account.
M. Shirakashi, K. Yaoeda, A. Fukushima, H. Abe; Effect of Optic Disc Size on the Diagnostic Ability of the Heidelberg Retina Tomograph in Japanese Subjects. Invest. Ophthalmol. Vis. Sci. 2009;50(13):5813.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To evaluate the effect of the optic disc size on the ability to discriminate between normal eyes and those with glaucoma in Japanese subjects when using the Heidelberg retina tomograph (HRT) glaucoma probability score (GPS) and the Moorfields regression analysis (MRA).
HRT (software version 22.214.171.124) images were obtained from a single eye in each of the 110 normal Japanese subjects and 136 Japanese glaucoma patients. In the glaucoma eyes, the average mean deviation (standard deviation) of the Humphrey full-threshold 30-2 program was -6.39 (6.42) dB. The samples were divided into three equal groups on the basis of disc size: small (disc area < 1.93 mm2, group I), medium (disc area 1.93-2.34 mm2, group II) and large (disc area > 2.34 mm2, group III). GPS and MRA diagnostic abilities in each group were evaluated by examining the sensitivity, specificity and area under the receiver operating characteristic curve (AUC), which were based on the overall classification results obtained when using a Japanese normative database.
When borderline cases were counted as test negatives, the sensitivity and specificity for GPS in groups I, II and III were 44% and 97%, 68% and 93%, and 88% and 61%, respectively, while for MRA, the values were 62% and 100%, 58% and 95%, and 82% and 84%, respectively. In both methods, larger discs were associated with higher sensitivity (P≤0.023) and lower specificity (P≤0.022). When borderline cases were counted as test positives, the sensitivity and specificity for GPS in groups I, II and III were 69% and 92%, 85% and 76%, and 96% and 42%, respectively, while for MRA, they were 78% and 87%, 78% and 83%, and 90% and 65%, respectively. For GPS, but not MRA, larger discs were associated with higher sensitivity (P=0.001) and lower specificity (P<0.001). The AUCs for GPS in groups I, II and III were 0.81, 0.86 and 0.76, respectively, while for MRA, they were 0.86, 0.83 and 0.85, respectively. There were no significant differences found between the AUCs among the 3 groups when using either the GPS or the MRA method.
For both GPS and MRA, disc size is significantly associated with the sensitivity and specificity of glaucoma detection in Japanese subjects.
This PDF is available to Subscribers Only