Comparison of Compass and Humphrey Perimeters in Detecting Glaucomatous Defects

Abstract

Purpose: To compare the performance of Compass fundus automated perimetry (FAP) and Humphrey Field Analyzer (HFA) in glaucoma patients.

Methods: A total of 120 patients with glaucoma underwent 1 FAP and 1 HFA perimetric test over the central 24° on one eye. The chosen eye and sequence were randomized and only reliable examinations were considered for analysis. Mean deviation (MD), pattern standard deviation (PSD), visual field index (VFI), and the area of absolute scotoma were compared between perimeters. Glaucoma Staging System (GSS2) data were analyzed by means of k test.

Results: Mean sensitivity difference (FAP-HFA) was -1.0 ± 2.81 dB (p<0.001, 95% confidence interval [CI] -1.61, -0.60 dB), MD difference was +0.27 ± 2.84 dB (p = 0.36, 95% CI -5.30, 5.83 dB), PSD difference was +0.48 ± 1.95 dB (p = 0.0075, 95% CI -3.37, 4.33 dB), and VFI difference was +2.4% ± 8.4% (p = 0.003, 95% CI -14.0%, +18.8% dB). Weighted kappa for GSS2 was 0.87. Points with null sensitivities were 9.9 ± 10.2 with FAP and 8.2 ± 8.9 with HFA (difference: 1.7 ± 4.0 points, p = 0.013).

Conclusions: Mean sensitivity with FAP is 1 dB lower than HFA, a finding due to different threshold strategies. Differences of global indices for FAP and HFA are small, which makes the 2 perimeters equivalent in the clinical setting. However, FAP seems more severe in evaluating glaucomatous damage, with absolute scotoma areas larger than with HFA. We raise the hypothesis that such difference may be the result of the active compensation of eye movements available with FAP.