Blumenthal EZ, Frenkel S.
Eye. 2005 Mar;19(3):308-11.
Department of Ophthalmology, Hadassah University Hospital, Jerusalem, Israel. eblumenthal@md.huji.ac.il
ABSTRACT:
OBJECTIVE: To evaluate the interdevice reproducibility of retinal nerve fibre layer (RNFL) thickness measurements obtained with the commercially available GDx-VCC, a scanning laser polarimeter with variable (individualized) corneal compensation.
METHODS: A prospective instrument validation study in which 13 GDx-VCC devices were tested. One eye each, from three normal subjects were used to test each of the devices, on the same day, by an experienced operator. Variability and reproducibility for each of five GDx parameters were calculated.
RESULTS: For each of five tested GDx parameters, the coefficient of variation and 95% confidence interval range (microm), for the 13 devices, respectively, were: TSNIT avg: 5.1%, 3.84 microm; Superior avg: 5.3%, 4.82 microm; Inferior avg: 6.1%, 5.50 microm; TSNIT standard deviation: 8.6%, 2.92 microm; and nerve fibre indicator (NFI): N/A, 5.69. Item reliability (Cronbach’s alpha) for the five GDx parameters are: TSNIT-Avg: 0.97, Sup-Avg: 1.00, Inf-Avg: 0.84, TSNIT-SD: 0.99, NFI: 0.99.
CONCLUSIONS: With the commercially available GDx-VCC, our results indicate that RNFL measurements appear reproducible across devices.
