Learning Effect of Short-wavelength Automated Perimetry in Patients With Ocular Hypertension

Abstract

To evaluate the learning effect of short-wavelength automated perimetry (SWAP) on a group of patients with ocular hypertension experienced with standard automated perimetry (SAP). Thirty patients with ocular hypertension underwent 5 full-threshold SWAP tests at intervals of 7+/-2 days. The parameters investigated to detect a learning effect were duration, the perimetric indices, and the number of points with a P of <5% and 1% in the total and pattern deviation maps. Differences in learning effect were also evaluated by comparing the sensitivities of central, paracentral, and peripheral areas, hemifields and quadrants. Learning effects were demonstrated for mean defect (P<0.0001, analysis of variance), duration (P=0.0001), the number of points with P<5% in the pattern deviation map (P=0.003), and short fluctuations (P=0.03). The effect was greater in the peripheral than in central areas (P=0.04). Mean defect was the most sensitive parameter, for which the learning effect was statistically significant between the first and the fifth test. The results of this study demonstrate a significant learning effect at full-threshold SWAP. This may limit the efficacy of this kind of perimetry in detecting early glaucoma, and should therefore be carefully considered when creating normative databases for new SWAP strategies.