Comparison of Methods for Visual Field Progression in Eyes with Central VF Defects

Abstract

Purpose : To investigate the consistency of various criteria for visual field (VF) progression in eyes with central VF defects.

Methods : Primary open angle glaucoma patients with 10-2 central VF defect (CVFD) at baseline with ≥2yr follow-up, and ≥5 visits for both 10-2 and 24-2 VF were included. 2766 24-2 VFs and 2797 10-2 VFs from 282 eyes of 197 patients were evaluated with the following methods: 10-2 clustered pointwise linear regression (PLR), 10-2 VF mean deviation (MD), 24-2 central VF mean total deviation (MTD), 24-2 VF MD, 24-2 PLR, 24-2 guided progression analysis (GPA), 24-2 Advanced Glaucoma Intervention Study (AGIS) and 24-2 Collaborative Initial Glaucoma Treatment Study (CIGTS) scores. Pairwise agreements among the methods were measured with Cohen’s kappa.

Results : The mean age of the patients at baseline was 68.9 (95%CI: 67.4, 70.3) years. Mean rate of 24-2 VF MD slope was −0.35 (95%CI: −0.35, -0.28) dB/y over 9.8 (95%CI: 9.3, 10.3) years, and mean rate of 10-2 VF MD slope was −0.33 (95%CI: −0.41, -0.26) dB/y over 9.7 (95%CI: 9.2, 10.2) years. In general, the proportion of eyes detected as progressing was between 20-35% for both 10-2 and 24-2 test strategies; fast progression was detected in a lower proportion of eyes. Proportion of series detected as progression was 20.6% for 10-2 clustered PLR, 35.1% for 10-2 VF MD, 17.7% for 24-2 central VF MTD (≤−0.5 dB/y), 3.2% for 24-2 central VF MTD (≤−1.0 dB/y), 19.5% for 24-2 VF MD, 20.6 % for 24-2 PLR, 30.5 % for 24-2 GPA, 17.7% for AGIS, and 27.0% for CIGTS. The kappa among the 24-2 methods was slight to 10-2 methods and 24-2 MTD methods slight to fair with ranges from 0.11-0.25, while the kappa between the 10-2 and 24-2 methods was fair to moderate with ranges from 0.22-0.54 (Figure 1).

Conclusions : The current study underscores the differences in agreement between visual field progression definitions based on comparing 24-2 VFs alone, comparing 10-2 VFs alone and comparing 24-2 and 10-2 VFs. The concordance between progression detection methods using the 24-2 VF and 10-2 VF is not consistently high, nor is concordance consistently high for different 24-2 VF methods for defining progression in eyes with central VF defects.

This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.

 

Figure 1

 

Figure 2.