To compare the abilities of standard automated perimetry
(SAP) and spectral-domain optical coherence tomography
(SDOCT) in diagnosing eyes with glaucomatous optic neuropathy
In a cross-sectional study, 280 eyes of 175 subjects referred to tertiary eye care center by general ophthalmologists for a glaucoma evaluation underwent retinal nerve fiber layer (RNFL) and ganglion cell complex (GCC) imaging with SDOCT. All subjects had at least 2 reliable and repeatable SAP. Two glaucoma experts masked to clinical and SAP results classified the optic nerves into GON and nonglaucomatous groups based on digital optic disc photographs. Ability of SDOCT parameters and SAP to discriminate GON eyes from nonglaucomatous eyes was evaluated using sensitivity, specificity, and likelihood ratios (LR).
Experts classified 179 eyes into GON and 101 eyes into nonglaucomatous group. Sensitivity of SAP (69.8%) was significantly lower (P
<0.05) than that of inferior quadrant RNFL thickness (83.8%), average RNFL thickness (77.2%), GCC focal loss volume (FLV, 81.5%), and global loss volume (GLV, 82.6%). Specificity and positive LR of SAP (95% and14.1, respectively) were significantly greater than those of all RNFL parameters, FLV (84.2% and 5.1) and GLV (82.2% and 4.6). Negative LR of SAP (0.32) was significantly inferior to that of inferior quadrant RNFL thickness (0.22), FLV (0.22), and GLV (0.21).
Most of the RNFL and GCC parameters of SDOCT had better sensitivities and negative LRs to diagnose GON compared with SAP. The specificities and positive LRs of most SDOCT parameters were inferior to that of SAP.