Purpose: To evaluate the reliability indices [fixation losses, false negative response rates (FN) and false positive response rates] and threshold sensitivities obtained from glaucoma patients with a Compass perimeter and to compare the same with the Humphrey field analyzer (HFA).
Methods: In a cross-sectional study, 97 eyes of 58 subjects (64 glaucoma and 33 glaucoma suspect eyes) underwent visual field examination with Compass and HFA. Any test with a fixation losses, FN or FP of >20% was considered unreliable. Reliability indices and threshold sensitivities between the 2 instruments were compared and the agreement evaluated using Bland and Altman analysis.
Results: In total, 37 tests (38%) with Compass and 17 (18%) with HFA were unreliable. The number of unreliable tests due to high FN (>20%) was significantly more (P=0.005) with Compass (n=27) than HFA (n=3). The mean difference [95% limits of agreement (LoA)] in mean sensitivity between Compass and HFA in the 51 eyes with reliable Compass and HFA results was -0.7 dB (-5.6, 4.3 dB). The point-wise threshold sensitivities with Compass were lower than that with HFA in central and temporal but higher in the nasal field. The 95% LoA ranged from -8 to +5 dB at one of the central points to -20 to +20 dB at one of the peripheral points.
Conclusions: The numbers of unreliable tests were higher with Compass compared with HFA. The LoA between Compass and HFA for point-wise threshold sensitivities as well as the global indices were wide, implying that the instruments cannot be used interchangeably.