Objective: To study prospectively the effects of perimetric learning in glaucoma patients.
Patients and methods: Twenty-five patients with newly detected glaucoma underwent repeated perimetric testing using the 30-2 full threshold program with the Humphrey perimeter. Each patient was tested five times at about 1-week intervals. No patient had undergone perimetry before entering the study.
Results: In most patients, visual field results improved with repeated testing. Improvements were obvious in 21 of the 37 glaucomatous eyes. Means of mean deviation values improved significantly, by 2.81 dB (P < .001, analysis of variance), between the first and the second test session, and no significant differences were shown between tests 2 and 5. Fields with moderate field loss improved more than fields with milder or more severe loss. Learning effects were more pronounced peripherally than centrally, and better points improved more than more disturbed ones.
Conclusions: The effects of perimetric learning were large and common, and are clinically important. Baselines for perimetric follow-up of patients with glaucoma should consist of more than one test. To avoid misleading conclusions, studies using perimetry to assess the effects of glaucoma treatment should take learning effects into consideration.