Objective: To compare the performance of 10-2 versus 24-2 visual fields (VFs) in detecting progression of initial parafoveal scotoma (IPFS) in glaucomatous eyes. Methods: Retrospective, observational study. Methods: Glaucoma patients with the following criteria: (1) an IPFS (≥ 3 adjacent points with P<0.05 within the central 10° degrees of fixation, 1 point or more with P<0.01 lying at the innermost paracentral points, and no scotoma outside the central 10°) in either hemifield based on 2 reliable Humphrey 24-2 Swedish interactive threshold algorithm standard VFs, and (2) 5 or more 10-2 and 24-2 VFs. Methods: Based on threshold map sensitivities, VF progression, defined as having 1 or more significantly progressing point(s) with a slope of sensitivity of less than -1.0 dB/year at P<0.01, was evaluated using pointwise linear regression. Methods: The number of progressing eyes in 10-2 and 24-2 VF analyses. Results: Fifty eyes (50 patients) were included (mean age ± standard deviation, 62 ± 9 years). Mean follow-up period (5.7 vs. 5.6 years) and number of VFs (7.6 vs. 7.8) were similar between 10-2 and 24-2 analyses (all P>0.3). Significantly more progressing eyes were detected in 10-2 than in 24-2 analyses (24 vs. 11 eyes; P = 0.007). This difference became greater within the central 10° (24 vs. 4 eyes; P<0.001). Four of the 11 progressing eyes in 24-2 analysis were missed in 10-2 analysis, whereas 17 of the 24 progressing eyes in 10-2 analysis were missed in 24-2 analysis. The 4 progressing eyes missed in 10-2 analysis had progressing point(s) only outside the central 10° in 24-2 analysis. The other 3 eyes with progressing point(s) only outside the central 10° in 24-2 analysis were detected as progressing in 10-2 analysis. Similar results were obtained when more stringent criteria (at least 2 significantly progressing points within the same hemifield) were used for VF progression. Conclusions: The 10-2 VF detects more progressing eyes than the 24-2 VF in glaucoma patients with IPFS, suggesting that closer surveillance of the central VF using testing algorithms with closely spaced grids is warranted in eyes with parafoveal scotomas.

Background: Proprietary or commercial disclosure may be found after the references.