Threshold Automated Perimetry of the Full Visual Field in Patients With Glaucoma With Mild Visual Loss.

Abstract

The authors used the Open Perimetry Interface to design a static automated perimetry test of the full field. Abnormal test locations in the nasal midperiphery and temporal inferior sector area best separated glaucomas from normals. The peripheral visual field in glaucoma outside 30 degrees is largely unexplored with static perimetry. Their goal was to use threshold static automated perimetry to characterize the visual loss in glaucoma of the central 30 degrees and the far periphery. The authors administered the 30-2 perimetric test to 27 patients with early stage glaucoma (with mean deviation better than -4 dB) with the Goldmann III and V stimulus sizes and a custom test from 30 to up to 87 degrees with the size V stimulus twice within a month. The authors quantified (1) the retest variability, (2) the proportion of patients flagged as abnormal (at level 0.05) on the basis of pointwise probability distributions obtained from 63 ocular healthy observers, (3) the pointwise statistical distance using the Kullback-Leibler divergence between normal and glaucoma eyes, and (4) the effect of eccentricity on visual loss. Size V 30-2 testing identified significantly more abnormal test locations (36%) than size III 30-2 (30%; P=0.004). Kullback-Leibler divergence between healthy and glaucoma distributions was greatest for the nasal midperipheral test locations and the inferior temporal sector area. A more pronounced decrease was found in visual sensitivity with eccentricity in the patients with glaucoma compared with the ocular healthy participants across the full visual field (P<0.001). Patients with glaucoma demonstrate a systematic decrease in sensitivity with eccentricity across the full visual field. Goldmann size V stimuli better detected visual loss in patients with glaucoma with mild loss than size III.