Purpose:
To evaluate the performance of scanning laser polarimetry (SLP) with variable corneal compensation (GDxVCC) to identify glaucoma in a population-based setting.
Methods:
1952 (76% of survivors) Blue Mountains Eye Study participants were reexamined after a 10-years follow-up. Routine eye examination including visual field (VF) examination and retinal nerve fiber layer (RNFL) measurement by GDxVCC were applied. Glaucoma was diagnosed when glaucomatous VF loss was present with matching neural rim loss on stereo photography.
Results:
A total of 1625 participants (3231 eyes) with the age of 73.5±7.6 years underwent SLP examinations. Typical scan score (TSS) ≥ 80 was found in 812 eyes (50.0%), or in both eyes of 516 subjects (31.8%). A total of 460 (28.3%) participants had reliable SLP imaging defined as both quality score ≥ 8 and TSS ≥ 80 in both eyes. The mean TSS decreased with: older age, lower quality score, higher TSNIT value, lower nerve fiber indicator (NFI), lower VF mean deviation (MD), and higher pattern standard deviation (PSD) (P<0.001), but not with gender (P=0.274), using a multivariate regression. The agreement of RNFL loss determined by GDxVCC, VF defect determined by MD, and the clinical diagnosis of glaucoma is shown in Venn diagram.At the NFI cutoff value of ≥30, GDx VCC had a sensitivity/specificity to identify glaucoma of 55.9%/82.5% in all participants, and 84.0%/88.0% in those with reliable SLP measurements.
Conclusions:
The proportion of subjects with low TSS in GDxVCC images was greater than previously reported. The high proportion of atypical scans suggests the GDxVCC would be of little value for glaucoma case-finding in an older, predominantly white, population. A new GDx device designed to improve TSS has been introduced.
Keywords: imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound) • clinical (human) or epidemiologic studies: systems/equipment/techniques • visual fields