Learning effect in automatic static perimetry with Dicon TKS 4000 computed perimeter

To find out the significance of the learning effect in repeated computed perimetry. We performed computed perimetry twice within one week period on 35 eyes (20 patients) with suspected glaucoma and visual acuity > 0.5 (20/40). The program used was with 30 degrees threshold grid. For each quadrant of the visual field were recorded the mean deviation readings. These readings were compared for the initial and the second perimeters for each eye. The mean deviation from the normal readings were significantly higher for the initial perimetric exam for all quadrants (Wilcoxon test for matched pairs, p < 0.05). The visual field defects, if present, were more concentrated and with sharper borders in the second exams. This study suggests that there is a significant learning effect in computed perimetry. A visual field exam, showing defects suspicious for glaucoma, should be repeated within one week and the second one taken for further referral.