Abstract
Purpose :
The size V stimulus has been found to have a greater dynamic range and improved repeatability [1] and is able to efficiently detect early glaucoma. We compared the diagnostic performance of a visual field test (VF) using a reduced number of test locations and the size V stimulus with the standard 10-2 test pattern.
Methods :
VFs using the Full Threshold strategy were acquired on a HFA3 Model 840 perimeter (ZEISS, Dublin, CA) for three combinations of stimulus size and test pattern: size V, reduced (32 point) pattern; size III, reference (68 point) pattern; and size V, reference pattern (Fig 1). While VFs were acquired on two visits, only data from visit 1 were used for further analysis. Size III and V reference limits were interpolated from a previous study [2]. Mean deviations (MD) and fraction of abnormal test locations flagged at 5% level for total deviation (TD) were computed. Comparisons between the reduced and reference patterns were carried out using Bland-Altman 95% limits of agreement (LOA).
Results :
The mean ages were 52.3 years (±14.5; 27.7 – 79.2; n=32) for the healthy group and 77.7 years (±5.3; 63.2 – 87.6; n=18) for those with glaucoma. Mean test duration (in minutes) was significantly shorter for the size V, reduced pattern (6.1±1.1) compared with the reference patterns using either the size III (11.1±1.7) and size V (11.1±1.9). MDs for the reduced pattern showed strong agreement with the reference patterns although slightly stronger for size V (r=0.98, LOA=-1.6, 2.3dB) than size III (r=0.97, LOA=-2.5, 4.1dB). Furthermore, there was a strong agreement between the reduced pattern and the reference patterns for both the size III (r=0.94, LOA=-0.21, 0.17) and size V (r=0.94, LOA=-0.22,0.16) in terms of the fraction of flagged points at 5% TD level (Fig 2). Visual inspection revealed good concordance between tests for the location of flagged points, with similar defects generally being present across all tests (Fig 1).
Conclusions :
Our preliminary findings suggest similar diagnostic performance between the size V reduced pattern and the reference patterns. Future work is required to compare these tests in terms of the similarity of locations for flagged defects and the ability to detect clinically relevant central scotomas.
[1] Wall et al. Arch Ophthalmol 2010 128(5):570-6
[2] Flanagan et al. IOVS 2016; 57(12).
This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.