Purpose: To highlight the learning effect in automated perimetry and to assess the value of the first visual field in glaucoma diagnosis.
Method: A prospective study that included 48 patients (91 eyes)--glaucoma suspects--who performed 3 visual field tests in no more than 12 months.
Results: 32 eyes were diagnosed with intraocular hypertension and the others with POAG. An improvement (expression of the learning effect) was noted in 45.05% of eyes (in an average of 3.39 2,72 points) in the second visual field and in 37.36% of eyes (3.05 2.13 points) in the third visual field. Considering (as a hypothesis) the third visual field as gold standard, the first test had a sensitivity of 87.03% and a specificity of 89.18% in detecting visual field defects (efficiency 89,01%). The second test had a sensitivity of 92.59% and a specificity of 89.18% (efficiency 91.2%).
Conclusions: The learning effect was present in the first three tests, but seems to have little influence on the diagnosis of visual field defects. A possible explanation is the presence to a large extent of important visual field loss (obvious from the first test) in the included patients.