Abstract
Purpose :
The Goldmann size V stimulus possesses better test-retest properties, identifies more defect in moderate glaucoma, and has a greater dynamic range than size III. The purpose of this study was to investigate the influence of large perimetric stimulus size on defect detectability in patients with early glaucoma, in particular for the most central stimulus locations.
Methods :
115 healthy (H; 71 female; 41 to 79 years) participants and 89 patients with early glaucoma (EG; 42 female; 43 to 82 years; average mean deviation (MD): -1.37 ±1.40 dB) were recruited from 3 sites. Subjects were included into the EG group if they had a clinical diagnosis of glaucoma, OCT findings characteristic of glaucoma, and a visual field (VF) MD of -4.0 dB or better (SITA-Std, 24-2). For 1 eye of each subject, 3 VF tests were obtained using the Full Threshold 24-2 (FT) for stimulus size III, V and VI; the VF tests were repeated within 90 days. Visit 2 data were used throughout. Data from the healthy group were used to produce reference limits for normality for each stimulus size.
Results :
See Table 1. Mean sensitivity (MS) increased with stimulus size and decreased with age. The variance for normal measures was lowest for size V, generating the smallest reference intervals. Size V identified the largest number of significantly depressed locations on total deviation (TD) analysis (~+1 location @ p<0.05), although there was no significant difference between sizes V and III. Size VI detected fewer TD locations (p<0.001). There was no difference in the number of TD flagged locations between stimulus sizes within the central 16 locations, with the differences being outside the central locations. Results were similar for pattern deviation (PD). ROC analysis of TD and PD demonstrated a significant difference between Size V and VI (p<0.02; p<0.05), but none between size III and V.
Conclusions :
Goldmann stimulus size V and III were comparable in identifying defects in early glaucoma. There was no difference in performance within the most central 16 locations of the 24-2, but size V flagged more TD p<0.05 locations outside of the central region. Size V generated tighter reference intervals for normality and lower test-retest variability. There was no advantage in using a size VI.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.