Agreement between Compass Fundus Perimeter New Grid and 10-2 Testing Protocols for Detecting Central Visual Field Defects

Abstract

Purpose : To evaluate the agreement between Compass New Grid (NG) and Compass10-2 test protocols for detecting early glaucomatous defects in the central 10 degrees of the visual field (CVFD).

Methods : 123 eyes of 14 healthy individuals, 17 glaucoma suspects and 32 glaucoma patients performed NG and 10-2 automated perimetry measurements within one week and macular scans using Spectralis OCT within 12 months. For both test protocols total deviation (TD) or pattern deviation (PD) plot CVFDs were defined in either superior or inferior hemifields by three contiguous points with probabilities of <5%, <2%, <2% or <5%, <1%, <1%. Cohen’s Kappa (k) statistic was used to assess agreement between NG and 10-2 for identifying CVFDs. The Spectralis GMPE Hood Glaucoma Report (investigational software version) macula deviation analysis was used to identify macular defects as areas of thickness with p ≤ 0.05 covering three or more contiguous 10-2 test points in the superior and inferior hemiretina for calculating sensitivities and specificities of all test protocols.

Results : Low to moderate agreement was observed between NG and 10-2 protocols for detecting presence of superior CVFDs on TD (k=0.566) and PD (k=0.256) plots and for detecting inferior CVFDs on TD (0.487) and PD (0.272) plots. Specificity of NG and 10-2 TD plots for detecting superior CVFDs was 0.82 and 0.65, respectively. Specificity of NG and 10-2 TD plots for detecting inferior CVFDs was 0.92 and 0.84, respectively. Specificity of NG and 10-2 PD plots for detecting superior CVFDs was 0.81 and 0.36, respectively. Specificity of NG and 10-2 PD plots for detecting inferior CVFDs was 0.86 and 0.52, respectively. Sensitivity of NG and 10-2 plots ranged from 0.20 for NG PD plot inferior defects to 0.78 for 10-2 PD plot superior defects.

Conclusions : Because agreement between the detection of central visual fields defects using Compass New Grid and 10-2 is fair to moderate, the visual field test results may be complementary and are not interchangeable.

This is a 2021 ARVO Annual Meeting abstract.