Background/aims A recent study has shown that the paracentral upper visual field in the macular region is often affected in glaucoma and suggested that two test locations within the central 10° should be added to the Humphrey 24-2 visual field test pattern to detect such damage. This study employed data collected using a different visual field test pattern to determine whether the same two-test locations are supported as the most informative regarding visual field loss.

Methods A data set of 62 patients with glaucoma and 48 controls had visual field assessments on the Medmont perimeter M700 (Central Threshold or Glaucoma test). Twelve 24-2 locations within central 10° of visual field were derived by interpolation of the nearest neighbours of the Medmont data. The remaining 24 Medmont locations in the central 10° of the glaucomatous set were labelled as abnormal if their thresholds fell outside the lower 5th centile of the age-corrected values for the same location from the control group. All possible pairs of the 24 locations were then assessed for diagnostic power by counting the number of patients that had 0, 1 or 2 abnormal locations in a pair.

Results Overwhelmingly, pairs of locations in the superior macular region were more often abnormal than pairs in the inferior. About 50 pairs of locations had equivalent ability to detect damage, with the best pair having 74% of patients with at least one of the locations as abnormal, and 52% both.

Conclusions Adding a pair of locations to the superior macular region of the Humphrey Visual Field 24-2 pattern increases the number of abnormal locations identified in individuals with glaucoma.